[Mrtrix-discussion] SIFT: normalise to the b0 image

Tue Dec 16 19:07:13 PST 2014

Hi Romain,
Yes, that is correct, the first (ideal) approach would be to divide all DW
and B=0 images by the mean CSF. For a recent AD study I actually used the
95th percentile of voxels within a CSF mask. Just to make sure I was not
getting any PV with grey matter.

As Donald said, at typically DWI resolutions this can only be done with
elderly populations. For cohorts below 60 years I normally use the median
b=0 WM value.  As Luis suggested, you could also use the average DWI
signal. However, if your WM mask includes voxels affected by pathology (eg
a reduced DW signal), then your normalisation will be biased and you may
remove some of the effect you are looking for. The b=0 intensity can also
be affected by disease and may bias your normalisation, however our
assumption is that pathology induced changes to T2 are less severe than the
DW signal. As long as the pathology is not global, then using the median
value for normalisation means you are less likely to introduce a bias. When
computing the median you could also mask out WM regions you suspect are
affected by the disease. A good sanity check to ensure that you are not
introducing a bias is to compare the median WM value across patient groups
with a t-test (to make sure there is no difference).

We normally use N4 to estimate the bias field on the b=0 and apply it to
the DW volumes. This seems to work OK, however we don't have really severe
bias fields.

Just to follow up on what Donald mentioned regarding problems with dividing
by the voxel b=0 in voxels with CSF partial volume. This is not just an
issue for spherical deconvolution and Apparent Fibre Density, but also for
tensor FA and ADC (see http://www.ncbi.nlm.nih.gov/pubmed/11323803).

Cheers,
Dave

On 17 December 2014 at 02:23, Luis Concha <lconcha at unam.mx> wrote:
>
> I have used fsl's FAST to estimate the bias field on the b=0 and applied
> it to the DWIs and it seems to work well.
>
> I guess one could also use some the segmentations provided by FAST to
> obtain the mean signal of WM voxels (of the average DWI) and use that to
> normalise the DWI signal between subjects.
>
> Dr. Luis Concha
> Instituto de Neurobiología
> Laboratorio C-13
> UNAM, Campus Juriquilla
> Boulervard Juriquilla 3001
> Juriquilla, Querétaro.
> C.P. 76230
> México
> Tel (442) 2 38 10 54
> Fax (442) 2 38 10 46
> http://personal.inb.unam.mx/lconcha/
>
> On Tue, Dec 16, 2014 at 8:41 AM, romain valabregue <
> romain.valabregue at upmc.fr> wrote:
>>
>>  Hello
>>
>> This is an interesting thread (worth to put on the wiki ... since few
>> people know that you do no take the B0 ...)
>>
>> Just to be sure I understand the first approach : you divide all DWI
>> volume by the mean csf value taken from the b=0 images ?
>>
>> What about about bias field correction
>> The bias is a big concern to: it is very large with 32 channel coils
>>
>> A few input on this ?
>>
>> N4 on the B0 and apply it to the DWI ? does it works well.
>>
>> Many thanks
>>
>> Romain
>>
>>
>> Le 16/12/2014 14:00, J-Donald Tournier a écrit :
>>
>> Hi Luis,
>>
>>  Well spotted. That is one of the issues that we still don't have a
>> totally satisfactory answer for. Basically, we need an internal reference
>> intensity for each subject so that we can rescale their data to a
>> comparable scale. One approach was to use the CSF intensity, which works
>> fine when the ventricles are sufficiently large that you can get a clean
>> estimate of the CSF signal free from partial volume artefacts. While this
>> works in older subjects, it's difficult to do in younger cohorts. Another
>> approach that we use is to measure the b=0 signal within a conservative
>> white matter mask, and use the median or mean value as the reference
>> intensity. This is effect scales the data to a single, subject-specific,
>> global WM b=0 intensity, avoiding the issues introduced by voxel-wise
>> normalisation to the b=0 signal. In all cases, the primary concern is
>> whether the intensity used as a reference is affected by the condition
>> under investigation, and hence whether this step may either mask a genuine
>> effect or introduce an effect that is not present in the DW data itself. My
>> guess is that the latter approach (normalisation to a subject-specific
>> global WM b=0 signal) raises the fewest concerns since it's closest to what
>> most people would naively expect anyway (i.e. normalisation to the
>> voxel-wise b=0 signal).
>>
>>  Happy to hear your thoughts...
>>
>>  Cheers,
>> Donald.
>>
>>
>>
>> On 15 December 2014 at 23:11, Luis Concha <lconcha at unam.mx> wrote:
>>>
>>> This is a very interesting discussion, thanks Rob and Donald for the
>>> detailed answers.
>>>
>>>  Rob, you mention that for inter-subject studies you would need
>>> a) B1 field correction (presumably accomplished by N3 or N4 or similar
>>> tools), and
>>> b) inter-subject intensity normalisation.
>>>
>>>  If not using a b=0 singnal normalisation, how can we get signal units
>>> comparable between individuals?
>>>
>>>  Dr. Luis Concha
>>> Instituto de Neurobiología
>>> Laboratorio C-13
>>> UNAM, Campus Juriquilla
>>> Boulervard Juriquilla 3001
>>> Juriquilla, Querétaro.
>>> C.P. 76230
>>> México
>>> Tel (442) 2 38 10 54
>>> Fax (442) 2 38 10 46
>>> http://personal.inb.unam.mx/lconcha/
>>>
>>> On Mon, Dec 15, 2014 at 7:18 AM, J-Donald Tournier <jdtournier at gmail.com
>>> > wrote:
>>>
>>>> Hi Jan,
>>>>
>>>>  Just to expand a little further on what Rob said, the main reason for
>>>> not normalising to the b=0 signal is essentially to preserve the apparent
>>>> fibre density. Really, the problem is that normalising to the b=0 signal
>>>> breaks the linearity of the DWI signal to the FOD (irrespective of the log
>>>> transform), and that is something we think should be avoided if at all
>>>> possible. This is an issue that I wish I'd described explicitly in my
>>>> original 2004 spherical deconvolution paper (even back then, all the
>>>> processing was done on the raw signal)... As Rob mentioned, we'll try to
>>>> rectify this is a future paper, but for now, here's a brief description of
>>>> my reasons for this.
>>>>
>>>>  This is based on a fundamental aspect of spherical deconvolution and
>>>> mixture models in general: that the DWI signal scales *linearly* with
>>>> the amount of tissue present. While the simulations done by Dave Rafflelt
>>>> in the paper Rob mentioned do make the point very nicely, their purpose is
>>>> a lot more specific than is required for this argument. Basically, If a
>>>> voxel contains two fibre bundles, the signal you measure is the sum of the
>>>> signals for each bundle individually (at least, it's modelled as such).
>>>> This however does not necessarily hold for the signal *attenuation*,
>>>> since the b=0 signal is not uniform throughout the brain.
>>>>
>>>>  Consider for example voxels containing mixtures of WM & CSF. The b=0
>>>> signal for CSF is typically very high relative to WM (due to its long T2).
>>>> If half the voxel contains CSF, the other half WM, the b=0 signal for that
>>>> voxel would be essentially double what it would be for pure WM (assuming
>>>> CSF b=0 signal is ~3x that of WM). On the other hand, the DW signal for CSF
>>>> is small, and to all intents and purposes negligible at high b-values. This
>>>> means the *raw* DW signal would be what you would expect to measure
>>>> for a voxel containing half the volume of WM, but the signal attenuation
>>>> would be halved again (since the b=0 signal is double). So the apparent
>>>> relative volume fraction (fibre density) derived using signal
>>>> *attenuation* would be ~a quarter that of pure WM, while using the*
>>>> raw* DW signal would give you the correct answer: half that of pure
>>>> WM. If you care about being able to compare apparent fibre densities across
>>>> voxels in the presence of large variations in the b=0 signal between
>>>> different tissue types (i.e. as you would expect in the brain), you
>>>> shouldn't normalise to the b=0 image.
>>>>
>>>>  Note this isn't just about voxel-based analysis of apparent fibre
>>>> density or SIFT: this is also important for example during the tractography
>>>> itself, since the termination criteria are applied on the FOD amplitude
>>>> directly. It is also important for anything that involves consistent
>>>> scaling of the noise (e.g. bootstrap analysis), since normalising to the
>>>> b=0 will also introduce large and rapid spatial variations in the noise
>>>> characteristics of the data. There are many facets to this issue, and I
>>>> won't bother going into them in any more detail here - I'll leave that for
>>>> the future paper Rob mentioned. But in a nutshell, this is the reason
>>>> MRtrix has always operated on the raw DW signal, not its attenuated
>>>> version.
>>>>
>>>>  Hope this all makes sense.
>>>> Cheers,
>>>>
>>>>  Donald.
>>>>
>>>>
>>>>  On 15 December 2014 at 00:32, Robert Smith <robert.smith at florey.edu.au
>>>> > wrote:
>>>>
>>>>>  Hi Jan,
>>>>>
>>>>>  This is an important point, and one that we sometimes forget that we
>>>>> (as in, the MRtrix dev team) think about quite differently to others in
>>>>> Diffusion MR.
>>>>> We will draw attention to this issue in an upcoming publication, but
>>>>> I'll try to give a succinct explanation here.
>>>>>
>>>>>  Conventionally, the log-transform with respect to the b=0 image
>>>>> converts a signal amplitude to an apparent diffusion coefficient; nothing
>>>>> controversial here. However if you were to then apply a spherical
>>>>> deconvolution transform, the FOD amplitude along a particular direction
>>>>> would be proportional to the ADC of the fibre population oriented along
>>>>> that direction. This isn't particularly useful information; it doesn't tell
>>>>> us much about differences between fibre populations throughout the image,
>>>>> or indeed within a voxel.
>>>>>
>>>>>  Ideally what we actually want for a number of applications is the
>>>>> volume of each fibre population element, in all voxels throughout the
>>>>> image. Based on David Raffelt's early simulations
>>>>> <http://www.sciencedirect.com/science/article/pii/S1053811911012092>,
>>>>> it turns out that (under certain conditions) the radial component of the
>>>>> DWI signal amplitude is actually a pretty decent marker for intra-cellular
>>>>> volume. Therefore, by ignoring the b=0 images completely and just running
>>>>> SD on the raw DWI intensities, we get pretty useful biological information
>>>>> and interpretation from the FOD; we also conveniently bypass the issue of
>>>>> Gibbs ringing in the b=0 images. Caveat is that you need a uniform B1 field
>>>>> (i.e. intensity bias field correction); for applications like AFD you also
>>>>> need inter-subject intensity normalisation, but that's not necessarily a
>>>>> problem for SIFT depending on how you're using it.
>>>>>
>>>>>  That's all for now. Hope that clarifies why we choose to apply SD in
>>>>> this way; in fact, this approach dates all the way back to the original SD
>>>>> paper.
>>>>> Rob
>>>>>
>>>>>
>>>>> --
>>>>>
>>>>> *Robert Smith, Ph.D*
>>>>> Research Officer, Imaging Division
>>>>>
>>>>> The Florey Institute of Neuroscience and Mental Health
>>>>> Melbourne Brain Centre - Austin Campus
>>>>> 245 Burgundy Street
>>>>> Heidelberg Vic 3084
>>>>> Ph: +61 3 9035 7128 <%2B61%203%209035%207128>
>>>>> Fax: +61 3 9035 7301 <%2B61%203%209035%207301>
>>>>> www.florey.edu.au
>>>>>
>>>>> On Sat, Dec 13, 2014 at 1:37 AM, Jan Schreiber <schreiber at cbs.mpg.de>
>>>>> wrote:
>>>>>>
>>>>>> Dear MRtrix Team,
>>>>>>
>>>>>> thank you very much for this great software and for making it freely
>>>>>> available!
>>>>>>
>>>>>> In your publication "SIFT: Spherical-deconvolution informed filtering
>>>>>> of
>>>>>> tractograms" you state
>>>>>>
>>>>>> "The diffusion signal must not be normalised to the b = 0 image
>>>>>> intensity. This preserves the linearity of the spherical deconvolution
>>>>>> transform between the measured DW signal and the resulting FOD."
>>>>>>
>>>>>> Shouldn't we preserve the linearity of the spherical deconvolution
>>>>>> transform between the FOD and the DW _signal attenuation_ rather than
>>>>>> the DW _signal_?
>>>>>>
>>>>>> Thanks,
>>>>>> Jan
>>>>>>
>>>>>>
>>>>>  _______________________________________________
>>>>> Mrtrix-discussion mailing list
>>>>> Mrtrix-discussion at www.nitrc.org
>>>>> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>>>>>
>>>>>
>>>>
>>>>  --
>>>>  *Dr J-Donald Tournier (PhD)*
>>>>
>>>>  *Senior Lecturer, **Biomedical Engineering*
>>>>
>>>> *Division of Imaging Sciences & Biomedical Engineering King's College
>>>> London*
>>>>
>>>>
>>>> *A: Department of Perinatal Imaging & Health, 1st Floor South Wing, St
>>>> Thomas' Hospital, London. SE1 7EH *
>>>> *T: +44 (0)20 7188 7118 ext 53613
>>>> <%2B44%20%280%2920%207188%207118%20ext%2053613>*
>>>>  *W: http://www.kcl.ac.uk/medicine/research/divisions/imaging/departments/biomedengineering
>>>> <http://www.kcl.ac.uk/medicine/research/divisions/imaging/departments/biomedengineering>*
>>>>
>>>> _______________________________________________
>>>> Mrtrix-discussion mailing list
>>>> Mrtrix-discussion at www.nitrc.org
>>>> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>>>>
>>>>
>>>
>>> _______________________________________________
>>> Mrtrix-discussion mailing list
>>> Mrtrix-discussion at www.nitrc.org
>>> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>>>
>>>
>>
>>  --
>>  *Dr J-Donald Tournier (PhD)*
>>
>>  *Senior Lecturer, **Biomedical Engineering*
>>
>> *Division of Imaging Sciences & Biomedical Engineering King's College
>> London*
>>
>>
>> *A: Department of Perinatal Imaging & Health, 1st Floor South Wing, St
>> Thomas' Hospital, London. SE1 7EH *
>> *T: +44 (0)20 7188 7118 ext 53613*
>>  *W: http://www.kcl.ac.uk/medicine/research/divisions/imaging/departments/biomedengineering
>> <http://www.kcl.ac.uk/medicine/research/divisions/imaging/departments/biomedengineering>*
>>
>>
>> _______________________________________________
>> Mrtrix-discussion mailing listMrtrix-discussion at www.nitrc.orghttp://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>>
>>
>>
>> _______________________________________________
>> Mrtrix-discussion mailing list
>> Mrtrix-discussion at www.nitrc.org
>> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>>
>>
> _______________________________________________
> Mrtrix-discussion mailing list
> Mrtrix-discussion at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>
>

-- 
*David Raffelt (PhD)*
Post Doctoral Fellow

The Florey Institute of Neuroscience and Mental Health
Melbourne Brain Centre - Austin Campus
245 Burgundy Street
Heidelberg Vic 3084
Ph: +61 3 9035 7024
www.florey.edu.au
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20141217/5d2df6d4/attachment-0001.html>