Dear Anna.
You can use a flexible factorial design to create the design shown by
Alexander. You would create three factors:
* subject: independent, equal
* group: independent, unequal
* treatment: dependent, (un)equal
and select main effect of factor 1 (subject) and interaction of factors
2 and 3 (group and treatment).
Best regards,
Guillaume.
On 25/01/17 14:50, Alexander Hammers wrote:
> Dear Anna,
>
>
> Sounds familiar but I haven’t built such a model in SPM12 yet… I’d just
> try and see what the design matrix looks like.
>
> In SPM</=8, it used to require something called “The Full Monty” :-).
>
> Hope this helps,
>
> Alexander
>
>
>> On 25 Jan 2017, at 08:18, BARNES, Anna (UNIVERSITY COLLEGE LONDON
>> HOSPITALS NHS FOUNDATION TRUST) <[log in to unmask]
>> <mailto:[log in to unmask]>> wrote:
>>
>> Thanks Alex and do you know how to enter that in the SPM12 gui? I'm
>> guessing it's full factorial with 2 levels? One for group and one for
>> before and after treatment?
>>
>> Anna ��
>>
>> On 24 Jan 2017, at 20:21, Alexander Hammers
>> <[log in to unmask]
>> <mailto:[log in to unmask]>> wrote:
>>
>>> Dear Anna and John,
>>>
>>>
>>> I think that in your design placebo/drug is better treated as
>>> categorical and not as a covariate.
>>>
>>> Check out p. 1012 of Hammers A et al. Brain 2007 where we’ve gone
>>> into some detail to explain our design choice for a “double paired”
>>> study. I have also appended the design matrix (showing the 1 -1
>>> contrast; 1 -1 -1 1 equally makes sense); note how paired scans are
>>> kept together but global values accounted for individually.
>>>
>>> Hope this helps, see you soon,
>>>
>>> Alexander
>>>
>>> <PastedGraphic-1.pdf>
>>>
>>> -----------------------------------------
>>> Alexander Hammers, MD PhD
>>>
>>> Professor (Honorary Consultant) of Imaging and Neuroscience
>>> Head of PET Imaging Centre
>>> Deputy Head of Division
>>> Division of Imaging Sciences and Biomedical Engineering
>>> King's College London
>>> St Thomas' Hospital, London SE1 7EH
>>>
>>> Telephone+44-(0)20 7188 8364 (PA Laura Zappulla)
>>> [log in to unmask] <mailto:[log in to unmask]>
>>>
>>>
>>>> On 24 Jan 2017, at 17:11, BARNES, Anna (UNIVERSITY COLLEGE LONDON
>>>> HOSPITALS NHS FOUNDATION TRUST) <[log in to unmask]
>>>> <mailto:[log in to unmask]>> wrote:
>>>>
>>>> Dear All,
>>>>
>>>> I have a SPECT study (although it could also be a PET study), where
>>>> I’m struggling with the creation of the design matrix and contrasts.
>>>>
>>>> I have a series of patients who have had two scans: one pre, and the
>>>> other post therapy.
>>>>
>>>> The patients are in two arms: placebo and drug.
>>>>
>>>> I was hoping to use a paired t-test paradigm with a covariate of
>>>> placebo and drug so that I could:
>>>>
>>>> - Look at the change in the placebo arm
>>>> - Look at the change in the drug arm
>>>> - Look for any differences between the two arms.
>>>>
>>>> My strategy was to use SPM12 in the following manner:
>>>>
>>>> - Use paired t test model
>>>> - Load paired scans
>>>> - Define a covariate on a per patient basis for placebo and
>>>> drug
>>>>
>>>> My first problem came with defining the covariate vector where I had
>>>> to define placebo/drug on a per scan basis rather than a per patient
>>>> basis for the model to work.
>>>>
>>>> With the ‘per scan’ definition I then struggled to define the
>>>> contrasts of change pre/post placebo, change pre/post drug,
>>>> difference in trial arms.
>>>>
>>>> Is my problem with the covariate vector? Should I have used a
>>>> different paradigm model? E.g what used to be called "multi-group
>>>> conditions and covariates". Would I use the full factorial option
>>>> to define this?
>>>>
>>>> I’m also as a next step hoping to factor in other covariates such as
>>>> age and disease status (on a per patient basis), but I‘m struggling
>>>> with the first step in my analysis L
>>>>
>>>> Can anyone help? Please?
>>>>
>>>> Many thanks,
>>>>
>>>> John
>>>>
>>>>
>>>> John Dickson Ph.D.
>>>>
>>>> Institute of Nuclear Medicine, University College London Hospital
>>>>
>>>> p: 0203 447 0525 m: 0793 124 9525 f: 0207 637 0578
>>>>
>>>>
>>>> ********************************************************************************************************************
>>>>
>>>> This message may contain confidential information. If you are not
>>>> the intended recipient please inform the
>>>> sender that you have received the message in error before deleting it.
>>>> Please do not disclose, copy or distribute information in this
>>>> e-mail or take any action in reliance on its contents:
>>>> to do so is strictly prohibited and may be unlawful.
>>>>
>>>> Thank you for your co-operation.
>>>>
>>>> NHSmail is the secure email and directory service available for all
>>>> NHS staff in England and Scotland
>>>> NHSmail is approved for exchanging patient data and other sensitive
>>>> information with NHSmail and GSi recipients
>>>> NHSmail provides an email address for your career in the NHS and can
>>>> be accessed anywhere
>>>> For more information and to find out how you can switch, visit
>>>> www.nhsdigital.nhs.uk/nhsmail <http://www.nhsdigital.nhs.uk/nhsmail>
>>>>
>>>> ********************************************************************************************************************
>>>
>>
>>
>> ********************************************************************************************************************
>>
>> This message may contain confidential information. If you are not the
>> intended recipient please inform the
>> sender that you have received the message in error before deleting it.
>> Please do not disclose, copy or distribute information in this e-mail
>> or take any action in reliance on its contents:
>> to do so is strictly prohibited and may be unlawful.
>>
>> Thank you for your co-operation.
>>
>> NHSmail is the secure email and directory service available for all
>> NHS staff in England and Scotland
>> NHSmail is approved for exchanging patient data and other sensitive
>> information with NHSmail and GSi recipients
>> NHSmail provides an email address for your career in the NHS and can
>> be accessed anywhere
>> For more information and to find out how you can switch, visit
>> www.nhsdigital.nhs.uk/nhsmail <http://www.nhsdigital.nhs.uk/nhsmail>
>>
>> ********************************************************************************************************************
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|