Medical Imaging Physics - Science topic

Dr. Robert M Nishikawa has some slides from an AAPM presentation that explain this. You may want to start with this and explore his references to answer the questions you are asking.

Hello Recep Kurtulus ,

The statement by Gerhard Martens is not contradicted, and the statement by Nobuyuki Hamada is supported (with the caveat that CeO₂ is added to stabilize the color of the radiation shielding and radiation-resistant optical glasses) in the following book.

[1] Heinz G. Pfaender, (revised and expanded by) Hubert Schroeder; Schott Guide to Glass; Van Nostrand Reinhold Company; 1983; p. 146.

Regards,

Tom Cuff

From one of the best group in the field :

An ultrasound image has poor resolution because the speed of sound varies by less than 10% and the speckle noise but with the new technologies overcome this problem

regards

i found that MRI siemens standart protocols is out of good imaging. This is may be my fail in understanding but nothing helps. so I hope edx format will help me.

@Rajeswari Rajendran:

Have you got any idea? I am using the XML file included in each case. but I wonder, how to map all 4 read in one slice?

You answered your question. Yes, external magnetic fields affect MEG recordings and quite easily can make them useless. What MEG measures is extremely weak magnetic field induced from neuronal currents. Background (external) magnetic field is much stronger and it would be probably impossible to measure brain magnetic field if the MEG recording rooms were not magnetically shielded.

Regarding the second question, check out TMS (Transcranial magnetic stimulation).

It was also helpful for me. Thanks!!!

Hi Nargess,

Short answer to your question:

Yes, you can. You will have to find out what Perfusion technique it is to make the best use of it and also you will need to 'read' 'delay time' between these images for a particular slice to process it correctly.  

You will find answers to most of your questions here:

Dear Jae Won,

I completely agree to Jeromes answer. Mixing data from different scanners is always a dangerous undertaking, especially in your case. You have to add the scanner type as a factor to your statistical model (e.g. GLM). Then look what scanner main effects and interactions with other factors are revealed by your data analysis. I expect, that these effects are larger than your actual "effects of interest". The whole situation is exacerbated if your scanner factor correlates with any "controlled" factor (extreme example: patient group examined on scanner A, controls examined on scanner B).

In addition, I would like to note that MPRAGE images of 1.5 T and 3 T differ systematically in WM GM contrast. The GM WM contrast decreases with increasing field strength. 

All the best,

Michael

...or simply normalise by size

For medical image analysis you can use Wndchrm.

If 3D information is needed for the management of the impacted tooth CBCT is indicated (This conclusion can be drawn on conventional X-ray imaging and clinical examination (palpation)).

If a 4cmx4cm volume is made with 90kV, low tubecurrent (2 to 3 mA) and limited arch rotation (180 degrees 9 seconds) this exposure can be made safely in children (dose aruond 10 micro Sievert).

This volume will not be crisp and sharp but it will give you the information at the lowest dose (ALADA as low as diagnostically acceptable).

If on the other hand the exposure is made with a large volume and custom parameters the dose can be 10 to 40 times higher.

The EGSnrc is a friendly Monte Carlo code where you can find the phase space for several linear accelerators. So you need to use  BEAMnrc to generate the beam and Flurznrc to calculate the electron fluence.

I had misread the coordinates before and was confused. Thank you for the answers

One easy solution would be to use a dilation morphological operator. Dilate your image by one voxel, then substract the original image from your dilated image. The result is the boundary region. You can then multiply by 0.5 and sum with your original image to have the boundary region in gray:

ana = analyze75read('test.hdr');

dil = imdilate(ana, ones(3,3,3)); 

boundary = dil-ana;

output = 0.5*boundary + ana; %may need to convert to float, since it'll round up if you're using ints 

Matlab also has a tool for 2D images called bwmorph. If a slice-by-slice approach is acceptable, you can run:

ana = analyze75read('test.hdr');

slice = ana(:,:,30);

bwmorph(slice, 'remove')

And then just repeat per slice.

1. For measuring T1, when using MRI (rather than spectroscopy), a spoiled gradient echo sequence (ie FLASH) done with different flip angles (eg 20 and 40 degrees)  is more accurate than either inversion recovery with different  inversion times or a saturation recovery spin echo with different TR's.

2. For measuring T2, use a multi echo conventional spin echo with 4 to to 8 echoes (preferably with non slice selective 180 refocusing pulses)

In both cases use a single slice to avoid cross talk or off resonance effects.

Remember that the slice profile of an MRI slice varies somewhat with flip angle (particularly at higher flip angles), which affects measurement accuracy.

From my experience in the remote sensing field (satellite, airborne, and shipborne imaging of the earth surface) high frequencies generally means a change in dn levels over a few to several pixels; low frequencies are change over a large number of pixels (i.e., broad patterns in the image).  To me strong and weak edges seems to indicate high vs low contrast with the surrounding pixels, but still high frequencies because the change is over a few to several pixels.  I have a few publications under my RG profile related to edge enhancement, including using statistical information to automatically select the approximate kernel (window) size of the spatial filter to enhance edges within a given image.

Pat

go for coif1 else db2. The oscillatory nature of PPG or most of the biomedical signals best matches with these two mother wavelets. Refer: http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=7359311&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D7359311

The simplest phase contrast imaging technique is the propagation based imaging and it has a good potential in dose savings. I would suggest to start reading the literature. You would need a $100K minimum for the setup that includes a source, detector, phantoms. Our lab is actively doing research in pbi and grating based PCI.

There are other imaging methods that can incorporate thermography.  Magnetic Resonance Imaging (MRI) can be sensitized to thermal measures that can be made in 3D and penetrate tissue that is optically opaque.  

Your project sounds more like an surface scanning technique with an image fusion component.  If that is your goal, many papers have outlined how to incorporate two different wavelength images with different resolutions.

 Hi Lian,

the high signal in the occipital region is due to surface coil or to be more precise: due to the fact that back of the subjects head lies on the coil (therefore is very close to the receiving coil). The purpose of "prescan normalize" is exactly to prevent or minimize this effect. Therefore you could use this option (normal mode should be fine). 

Another option is to use a thicker cushion to raise the subjects head (and thereby achieve similar distances to all parts of the coil + more comfort for your subjects).

Nevertheless it's astonishing that you report systematically raised Z-scores in this region: by theory Z-scores should be independent of the magnitude of the signal because mainly (normalized) differences should influence the Z-scores. Probably increased SNR by the increased signal could be an explanation but this effect is not very strong.

Dear All,

The best is to use so called optimal pixelsize, where resolution and noise level can be optimized by relatively easy way.

        PSopt~FWH/3, where PSopt is the optimal pixelsize, FWH is the system (currently the SPECT system resolution including both the detector resolution and the selected collimator resolution). If you follow this approximation you can be optimized any SPECT study.

Hello,

You need to apply a bilinear transform to convert Hounsfield Units to linear attenuation coefficients. Some workstations have software licenses that allow you to use an external CT for attenuation correction (i.e.: Syngo from Siemens, Xeleris from GE, etc.). 

You can find the required equations is this paper:

I have a MATLAB routine that implements this conversion, please email me if you want to try it. However, you´ll need reconstruction software that allows you to input these mu-maps to correct for the attenuation. 

Best,

Mauro 

hi read this article i think will be useful for you

It turns out that the main contribution to T1 of brain tissue is myelin concentration. The most quantitative paper describing this so far is Stüber C, Morawski M, Schäfer A, Labadie C, Wähnert M, Leuze C, Streicher M, Barapatre N, Reimann K, Geyer S, Spemann D, Turner R. Myelin and iron concentration in the human brain: a quantitative study of MRI contrast. Neuroimage. 2014 Jun;93 Pt 1:95-106. doi: 10.1016/j.neuroimage.2014.02.026. Epub 2014 Mar 6. Since the myelin content is the main biochemical difference between grey and white matter, it makes sense to focus on myelin in trying to understand white matter disorders.

Because so-called T1-weighted images are also proton density and T2*-weighted, and they are not quantitatve, scientifically speaking there is little point in acquiring them. If one really wants to understand white matter pathology, one should use T1 mapping sequences, such as MP2RAGE, and quantitative proton density sequences, such as TAPIR. If fiber structure is of interest, one should use DWI approaches such as DSI, and quantitative susceptibility mapping, which reveals fiber orientations, iron concentrations and diamagnetic inclusions.  'White matter integrity' is a vacuous concept which should be abandoned except in the case of frank white matter disease, such as MS or ALS (see Jones D et al 2013).

I hope this helps.

Darrion, you can find here an example of pelvic cavity phantom: http://www.cirsinc.com/file/Products/002PRA/002PRA_DS_070113.pdf

I agree with the method provided above. But I wonder why in the first place the author got the t values and the coordinates in an xls file to start with. If this is fMRI imaging processing result, I suggest to use SPM to analyze the data and the final result (with clusters) would be easy to get, which would make your life easier. 

Yes of course there are some. Here below is one ref but you could find others easily.

Solid focal liver lesions indeterminate by contrast-enhanced CT or MR imaging: the added diagnostic value of contrast-enhanced ultrasound.

Quaia E.

Abdom Imaging. 2012 Aug;37(4):580-90.

Hi Carles

I am using stir mailing list but i did not found anyone there.

However i did some progress and used projection output (.hdr + .sin) for SPECT reconstruction in STIR but some problems remained. The key is true changing in header file that i have some problems with that.

Best

Thanks a lot for the answer !!

Armando, as you mention, neuroimaging is a very vast field and various concepts in mathematics (linear algebra, geometry, differential geometry, manifolds, numerical analysis, graph theory, etc.), statistics (multivariable analyses, advanced statistical learning and modeling, Bayesian modeling, etc.), mathematical physics, signal processing, neural networks, and others are used in various subfields of neuroimaging. For that reason, there is no real place to start unless you focus down on a specific aspect of neuroimaging. For one way of starting, you can look at some of the papers by Karl Friston and his colleagues, who is one of the most prolific writers in the area. His groups has a lot of appled papers but you can filter through and find the methods papers and take it from there. 

Lack of consensus on the voxel size reflects the statistical issue of multiple-tests in fMRI studies: the bigger the number of voxels the higher must be the adjustment, which leads to statistical significance evaporating very fast.  In my view, the way to resolve this issue with the present sample sizes is by defining a very specific and highly plausible null-hypothesis and then trying to falsify it.

Hi Amir,

there is no simple answer to you question since any wavelength interacts with tissue. So if you do not kwow which physical effect you want to study. Diffuse Reflectance is usually due to body scattering. Scattering itselfe is dependent on the wavelength and the size of the targets (and other things, maybe have a look at wikipedia or a textbook). 

You should also consider absoption. If you choose a wavelength that is, e.g. stongly absorbed by water it migth not give you much signal in refectance.

As for the difference between lasers and LED, laser are basically strictly monochromatic, means emitt light with one exact wavelength, LED emitt ligth with an band of wavelengths, means a maximum wavelength and bandwidth.

Hope that gives you some directions.

Our group has published extensively on breast SPECT with Sestamibi, including data on typical uptake values, count densities, etc. For example (see linked paper below), early results from a pilot clinical trial indicated that the average uptake of Sestamibi in healthy breast tissue is approximately 0.10uCi/mL (with a 25mCi initial injection). You could easily use this value with your given simulated breast size to compute the number of histories necessary (assuming isotropic emission).

Ok Dr. Gsell . You made an important observation . Let's wait for the combinations to see the extent of real gain in terms of diagnosis.

Find the center of your wire object and take ROI across it, that's a point spread function for you. Integrate along y direction will get the line spread function (LSF). Take the Fourier transform of LSF that's your MTF.

Beam hardening has a number of effects on dose and image quality, so a simple answer does not suffice.

When a beam hardenes, the relation between tissue density and attenuation changes too. A higher photon energy results in reduced densitiy discrimination.

When a beam hardens the average photon energy increases and the probability of that photon being absorbed decrease. So these photons are less likely to be absorbed and contribute to the dose.

Various beam modulation techniques are frequently used for dose reduction. This may interfere depending upon design.

In extreme cases [i.e. metal artifacts] both the beam hardening effect is quite large locally and locally detector noise becomes a serious problem. Hence the streak artifacts.

All of these effects interact in a non-linear fashion and also depend on the filtering used for the tube.  As manufacturers differ largely with respect to the amount of filtering with their x-ray tubes, I expect a considerable variation in the extent of the resulting image degradation due to beam hardening.

In addition the tube current may be an additional factor: in case of a low tube current, the noise floor of the detector system is likely to be reached sooner and hence the image deterioration under conditions with significant beam hardening become much more noticable.

In dual energy CT the effects are different for the low- and high energy beam.

In short: it depends on all of the above and on how the respective manufacturers try to compensate for the various effects.

Here is a quite comprehensive database of journals with listed IF:

The Cancer Imaging Archive might be worth a try.

There are several approaches to retrieve good images. Thirst question is what do you really need/expect? You won't get any SUV from PET/CT until you made your own calibrations and very time consuming scans. To get nice images we use 25-30 min per bed position (about 18 cm with Philips Gemini TF and Time of Flight) for patients. I know that Philips was working on a calibration factor so that you can easily choose Y-90 like F-18 or Ga-68 in the acquisition protocol. I'm not sure if they succeeded or if such a calibration factor is available from GE. Maybe you've to contact GE Service. SPECT/CT is well described in several papers. One method is to use medium energy collimators and a wide energy window (20-30%) @75 keV to use Bremsstrahlung for imaging. This works fine for our Siemens Symbia T6 with 60 Frames over 360 degree with 15s per frame, continuous acquisition mode and auto contour. Reconstruction is 3DOSEM 8 iterations and 8 subsets and no filtering. It is necessary to perform scatter and attenuation correction. Maybe you have to set the energy to 75 keV for scatter and attenuation correction as well in your recon protocoll. I want to make clear that this is for routine use in patients, not for quantifying activity concentrations. There you may have to add more frames and more time per frame and play a little with reconstruction parameters. Hope this helps.  

If you are willing to get into a bit of programming, you could try the vmtk libraries (http://www.vmtk.org/). The authors would probably mix in 3d Slicer for free registration and visualization. They seem now to have a commercial offering which presumably makes life simpler, but not free. 

We have done some work and are using allometric scaling a la Atkinson 2013 - seems to be the best way to analyse the dat if you are looking at pre- post- design

In my experience, in  The Netherlands, no consultancy firm exists, that combines knowledge of the operational size AND of the legal side AND of the radiation protection side. So it is extremily important to find a few experienced professionals [not just physicists, but also physicians] to monitor any such project on the sidelide.

Take for example the impact that the introduction of PET and PET/CT has had on the design / radiation protection requirements.

In 2004 I had the [College Bouw Ziekenhuis Voorzieningen] (Dutch Governmental Advisory committee for hospital building visit our place, because we judged that their rules and regulations were not compatible with modern radiation protection standards. [At that time our government had a large say in these matters, as the government financed hospital buildings. This system has been abandonded.]

Any rules and advises should be judged from the perspective of the time frame in which these were formulated. And most certainly, one should think ahead. When we finalized our plans for a new department in 2008, we based it on a prognosis for 2025.

I totally agree with Farshid Salehzani that architects are a risk factor rather than a source of good advice. I have dealt with three designs, 2 full departments and one PET/CT centre and know for certain, we have avoided a financial catastrophe each time, by insisting on full participation in the design loop.

In the beginning we were considered a nuisance. But once the concrete was poured for the foundation, we started inspections. This led to a number of costly findings, causing serious delays. Several things had to be redone. We explained our the board of directors that we would apply for our licens because we considered the findings so serious that it was not worth the effort of writing the application for a license. The board of directors acted promptly and since then we were timely  invited to comment on all major design issues and at various stages we were activily invited to inspect the details, i.e. of how the issue of shielding was solved arond the plumbing, piping and wiring. The constructor was very keen on giving us an opportunity to inspect BEFORE details were hidden by concrete whereever possible.

That no experienced architect can be found in most countries is no surprise. In The Netherlands, only 1 or 2 hospitals are built each year. Assume 5 consortia that build hospitals, no single architect can acquire the experience needed and at the same time keep up with the legal stuff and the radiation protection stuff and also maintaining an up to date overview of what contemporary nuclear medicine is about.

If you want spatial precision at the level of identification of cortical areas, there is no point in using VBM at all, which is intrinsically a fuzzy and non-quantitative technique that irretrievably confounds spatial extent and intensity. Much better to use maps of cortical thickness, derived from FreeSurfer or CBS_Tools. Registration of groups of brains using surface registration is also to be recommended, preferably with myelin maps. See Tardif et al 2015.

Your service contract should include periodic performance evaluations of your system and confirming your gradient slew-rate would be a part of this evaluation. If you really need to confirm it yourself, there should be sample ports for each axis in your gradient cabinet. You can connect these ports to a storage oscilloscope and record the gradient waveforms. Keep in mind though, that depending on the pulse sequence parameters your using the observed slew-rate might be limited to less than 200 T/m/s. 

Thanks, Yes it is a dosimetry phantom used in radiotherapy. We think it is a Varian

Your answer surprises me. With a dose of 80 MBq at rest and a normal collimator, ignoring decay / scatter / attenuation / part of patient outside FOV / geometric efficiency / poor algorithmic efficiecy, your 'true' data compromises at most 6E5 events amidst a lot of noise. Given that likely half the photons are scattered, half are attenuated and three quarters will be outside the field of view and that probably the GP will compromise less than 1.5% of the MIBG uptake within the field of view, I am not convinced that it is at all feasable. 

Hi Haleh, I cannot think of any existing picture norms have that information. The closest one that I think still quite helpful in your case comes from Cree and McRae (2003) and their following norm publication in 2005. In their study, they are people to write down features associated with >500 concepts across many categories. Based on the featuring listing data, they calculated the visual similarity measures for each subcategory (e.g., fruits, weapons, vehicles, birds, animals, etc.). Although these data comes from verbal modality, I think they may help you get a rough sense of the visual similarity for the subcategories. 

For collecting picture rating data, in order to get similarity data, the traditional method is to collect all pair-wise comparisons and calculated the rated similarity scores for each subcategory. This will be painful. I will suggest you using the adaptive sampling algorithm which is powerful and easy to use. I have add the links for relevant paper and website for the adaptive sampling.  Hopefully, it will be helpful! Good luck!

You may look at article "METRICS PERFORMANCE COMPARISON FOR COLOR IMAGE DATABASE. Nikolay Ponomarenko (*), Federica Battisti (**), Karen Egiazarian (***), Jaakko Astola (***) and Vladimir Lukin (*)

(*) National Aerospace University, Kharkov, Ukraine

(**) University of Rome "Roma TRE", Rome, Italy

(***) Tampere University of Technology, Tampere, Finland.

ABSTRACT

In this paper, we exploit a new database of distorted test

images TID2008 for verification of full-reference metrics

of image visual quality. A comparative analysis of

TID20008 and its nearest analog LIVE Database is

presented. For a wide variety of known metrics, their

correspondence to human visual system is evaluated. The

values of rank correlations of Spearman and Kendall with

the considered metrics and Mean Opinion Score (MOS)

obtained by exploiting TID2008 in experiments are

presented. The metrics are verified for both full set of

distorted test images in TID2008 (1700 distorted images,

17 types of distortions) and for particular subsets of

TID2008 that include distortions most important for

digital image processing applications.

Bernoulli Statistics still apply. The E-field is non-coherent, the energy density is still 1.0000(00). Homogenous distribution is an assumption, the Lagrange invariant still applies. An average Coherence length should still apply.

PBR28 compound of either [18F]fluoromethyl-PBR28 or [11C]PBR28 have been used to study Parkinson's disease. A recent study described its use to image a biomarker translocator protein (TSPO): Bioconjugate Chem., 2014, 25: 442-450.

There are several databases that come to mind for high-quality (and high-volume) data from normal subjects, some of these are listed in the NITRC, but there are likely others too:

1. (A collection of data from across the globe, aimed at fcMRI, but likely contains useful data for you) http://fcon_1000.projects.nitrc.org/

2. (Very high quality data and analysis tools under development) http://www.humanconnectome.org/

3. (The BIRN group has a list of databases) http://www.birncommunity.org/resources/data/

4. (The ADNI study of Alzheimer's contains normals too) http://adni.loni.usc.edu/methods/mri-analysis/adni-standardized-data/

5. (Other end of the spectrum, pediatric scans are available too) http://www.pediatricmri.nih.gov/nihpd/info/index.html

CBCT is the best method for evaluating the jaws for implant planning because you can evaluate the bone in 3 dimensions but in panoramic you can't evaluate the jaws  buccolingually. linear measurment in panoramic radiograph isn't  accurate but in CBCT is accurate. 

Thanks Mustafa.I want to know how to control speed.I mean vary the speed to known values.please let me know if you have any info.thanks

Hi,

Contrary to raw AC EEG datasets, the baseline about which the signal oscillates in raw DC EEG datasets actually changes over time (DC drift). You should not have any problem comparing them to each other as soon as you equally remove low frequencies in each other.

Be also careful when cleaning/removing artifacts such as eye movements as they can in certain cases take different shapes when recording in AC/DC.

Here is an article that you might find interesting : Hennighausen E, Heil M, Rösler F. A correction method for DC drift artifacts. Electroencephalogr Clin Neurophysiol. 1993

Regards.

The DTFIT tool in the FDT toolbox of FSL is what we use, then feed this back into MATLAB to work with the diffusion tensor.

Dear Gozde,

thanks so much that would be perfect, I will email him and read your paper with great interest. 

Will

Speckle noise is the primary factor affecting the appearance in ultrasound images or videos. This is a multiplicative noise which requires appropriate modelling and should be filtered out. There are a large number of methods proposed in the literature as well as appropriate software for ultrasound image/video despeckling. By preforming despeckling you also increase image contrast. Maybe you may look at our book:

C.P. Loizou, C.S. Pattichis, “Despeckle filtering algorithms and software for ultrasound imaging,” Synthesis Lectures on Algorithms and Software for Engineering, Ed. Morgan & Claypool Publishers, 1537 Fourth Street, Suite 228, San Rafael, CA 94901 USA, June 2008, ISBN-13: 9781598296204.

Alternatively, also have a look at research gate under my page. There are a lot recent publications made from our group accompanied by software (which can be downloaded) for image and video despeckling. See also our website at http://www.medinfo.cs.ucy.ac.cy/

There is very little that the number of components found could tell you about brain functioning or connectivity, since many of the components could be related to noise, artefacts or be vascular components. You should take some considerations into account. Did you perform group ICA with temporal concatenation? Did you use automatic estimation of the number of ICs?

If you really want to compare the integrity of the resting state networks in a group of patients vs. controls your approach is not adequate.  I suggest that you go for a entire sample group analysis, and after identification of the real resting state networks you apply a two-groups differences analysis. Take into account structural differences into account. In the end, differences found would reflect differences in network integrity/pattern. 

The CR systems uses a log amplifier. There is one advantage as this roughly corrects the signal to be proportional to the density of everything in the beam i.e. inverting the effect of the exponential absorption properties. This is useful to know if you are looking at a raw clinical image (with linear LUT), it can be easier to view after applying a log transform to the images.

There is more and I am stretching my brain for this one, but there are some advantages in terms of noise for log amplifiers. I think linear amplifiers are used for DR as log amplifiers would amplify the electronic noise in low signal regions - this is not an issue for CR as there is negligible electronic noise.

Either way the images are generally presented to the reader with a sigmoid look up table.

It makes no sense to use a "small volume" that includes parts of the image where brain activation cannot possibly occur--i.e. in white matter, bone or CSF. Thus the spherical "small volume" normally used in SPM is fundamentally inappropriate, since it would need to be less than 3 mm diameter to not include WM, bone or CSF. It is greatly preferable to use a shaped small volume that includes only grey matter, from one specified and well-defined cortical area, or one deep nucleus. This can only be achieved when a suitable individual-brain cortical and deep brain parcellation has been performed. See my recent paper 'Comparing like with like: the power of knowing where you are' (Turner R, Geyer S. Brain Connect. 2014 Sep;4(7):547-57) to see further details of this approach, which takes advantage of the unique sensitivitiy of MRI to details of myeloarchitecture and iron distribution. To use such small volumes considerably increases the experimental power, and enables well-grounded averaging across subjects and comparison between subject groups. Once this approach has been widely adopted, we will be able to place far more confidence in fMRI findings, in any type of study

As I always say in this kind of opened and general questions you must be more clear and define better your problem in order to ask something that really helps you. According to the way your question is formulated you can get thousands of answers, all of them probably right, but also probably useless for your problem.

You should clarify basic information like:

Once you give such details, it will be easier to help you. Otherwise, from my point of view, any suggestion makes no sense. :)

Cheers.

Bruno,

Some older studies comparing 15O water PET and fMRI mainly showed the same regions but with different centers of mass. This has been attributed to the difference in the methodologies. PET measures 150 Water as it is perfused into the areas of the brain that are active while fMRI measures the oxygen depleted hemoglobin in the blood after it has been extracted by the active cells. Depending on the area of activation, the directional flow of blood, and the spatial resolution of the imaging technique, the difference in the centers of mass has been roughly one centimeter. 

More information can be found in this study:

Hi,

yes, the crushers have no effect on the transverse magnetisation if you place a one crusher before and after the refocusing pulse - given they are of similar strength. If you include molecular diffusion you will get an additional damping of the signal - which is used to measure ADCs in NMR.

By looking at your MATLAB code it seems to me that the offset term is not implemented correctly. I was just loading a 90˚ hard pulse of 1ms and saw that it gives perfect excitation for +- 10 kHz which cannot be true. Maybe check again, how the offset term is implemented, it should lead to a nutation angle for off-resonant excitation.

Also, I can recommend this paper here for selective refocusing pulses:

Best regards,

Franz

I want talk about HIFU monitoring with MR and ultrasound. Compared with MR, ultrasound has better temporal resolution and lower cost. However, currently MR thermometry is the only clinically-feasible temperature imaging method. Ultrasound temperature imaging techniques are under heavy investigation, but still far from being clinically applicable. This is mainly due to the low temperature estimation accuracy and limited applicable temperature range (most of them are applicable in 37-45 degrees Celsius). Another aspect of ultrasound-based HIFU monitoring is related to thermal lesions (or ablation zones) when tissue coagulation necrosis forms. This aspect is under extensive exploration too.

Dear Amir.

Check the following references that you can download from my profile. Both relate to different physical phenomena that contributes to the final PET point spread function.

The influence of photon depth of interaction and non-collinear spread of annihilation photons on PET image spatial resolution. Eur J Nucl Med Mol Imaging. 2006 Aug;33(8):940-7.

Positron flight in human tissues and its influence

on PET image spatial resolution. European Journal of Nuclear Medicine and Molecular Imaging Vol. 31, No. 1, January 2004

There can be a lot of factors, I would add to the list movement artefacts and metal objects in the region of interest.

There are a few ways with which you may compare similarity of images:

The stuctural similarity index (SSIN) is a good method proposed by Bovik et al. There are also other traditional metrics such the MSE, SNR, PSNR and many more.

You may find a complete analysis of all these metrics in a recent publication from our group:

C.P. Loizou, C. Theofanous, M. Pantziaris, T. Kasparis, “Despeckle filtering software toolbox for ultrasound imaging of the common carotid artery”, Comput. Meth. & Progr. Biomed., vol. 114, pp. 109-124, 2014. Best paper, editor’s choice.

There is also a software available to download with which you can perform analysis and get similarity meausres between images. You may download this from the researchgate.

Dear Bashir Kayode :

This is an exciting and useful problem. Your question is very general and is complex to provide a satisfactory answer. Maybe you should elaborate a bit more on your problem.

There are several methods to obtain maps of T1 and T2. But more than a tool of Matlab , you need to design and select your experiment. I recommend taking into account, among other things:

1 - Pulse Sequence . The equations describing the variation of the intensity of a pixel is different in each case .

Spin Echo ( SE)

a) Calculation of T1: TR change with a fixed TE .

b ) Calculation of T2: TE change with a fixed TR .

IR pulse sequence :

a) Calculation of T1: IT change and fixed TR and TE .

Gradient Echo (GRE) . Change the FA and hold fixed TR and TE .

2 - Select the calculation method : generally using optimization methods ( Levenberg-Marquardt ) .

Anyway , your question has various answers depending on many things. I have some Matlab functions for T1 and T2 map calculation using the DICOM images. I can give you is the if you need them.

If you want more help you can contact my mail: evelio.gonzalez @ cigb.edu.cu

I am enclosing an article that can guide you solve your problem .

Best regards .

No difference if they are traveling at the same speed. May be the signal strngth could be different.

No problem with your understanding about the focal length.

However when I say 'fan-beam' geometry, it already means a 2D-to-1D implementation; its 3D-to-2D counterpart can be 'cone-beam' geometry.

Pinhole geometry is something else usually used in emission imaging instead of transmission imaging. A pinhole is a very small hole through a shielding material between the object and the detector, not a source like the focus spot in a X-ray tube. See a drawing of pinhole geometry here: https://en.wikipedia.org/wiki/File:Pinhole-camera.svg

Hi...I am not sure if it will help you a bit....However, I can suggest you two websites with some US images of the breast, though they do not provide a real database: http://www.medison.ru/uzi/eng/all/mammography.htm

Absorbed dose to an organ is defined as the energy absorbed (from radiation) divided by the mass of the organ.

In order to calculate the energy absorbed in an organ, you need to know the total activity of the pharmaceutical in that organ and such quantity is time dependent.

You need some information such as the bio-kinetic model (biologic distribution) for the animal that you have used in your experiment. You also need organ dose calculation software that will sum up all the radiation energies from all the source organs to the target organs in the animal under study.

The most common software is known as OLINDA (organ level internal dose assessment) written by Dr. Michael Stabin, of Vanderbilt University; but this software is applied to human data and not to animal data as far as I know! But you can check out the software applications available online. I guess you may be able to benefit from the software phantom results by using the following relation

A (phantom) = A(animal) [( m/Mphantom) /(m/Manimal)] where m is the organ mass and M the total body mass.

using the animal organs mass to obtain an approximation of the doses to the organs; anyway what is done is usually an estimate of the organ dose; do not forget that an estimate has always an associated uncertainty.

To convert the measured counts in each organ to an activity in (Bq) deposed into the organ; you need to calculate a factor relating the counts registered by the gamma counter used and the corresponding standard radiation source (radiopharmaceutical) activity in (Bq). Then once you have the determined the activity you need another factor that will relate the integrated Activity in time measured in (Bq.hr) to the absorbed dose in (rads) and this is what is missing the so called (S) factor. You need the software for that because (the software uses Monte Carlo simulations results and mathematical phantoms).

Good luck

Hi,

by superparamagnetic and 50 nm size you mean "hydrodynamic" diameter, right?

The point is a big amount of magnetic material within a localized region may preclude you using standard medical protocols for MRI. But it does not mean you cannot employ another apporaches.

Anyway, I cannot help you about the design of the device (let's hope somebody else in the forum can help you). Regarding the nanoparticle agent, I would suggest you (both for MRI and also SAR optimization) to use iron oxides about 40 nm (crystal core) -obviously not superparamagentic at RT in quasistatic conditions. You may find a nice survey on the issue by Quoc L. Vuong et al, DOI: 10.1002/adhm.201200078.

of course, aggregation of the particles, their coating and geometry of the assembly -if any- would also modify their response.

Best.

Great can we exchange emails to start the process

Hello. I recommend you compare their results on the same DICOM image with those obtained with some professional software, compare the results of the areas and volumes. The Amira 5.x is very good for this. You can also use a phantom with several extructura and segment with its algorithm and the Amira and compare them.