Dear Ashkan,

T2*-weighted sequences are used to depict paramagnetic deoxyhemoglobin, methemoglobin, or hemosiderin in lesions and tissues. Pathologic conditions that can be depicted with these sequences include cerebral hemorrhage, arteriovenous malformation, cavernoma, hemorrhage in tumor, punctate foci of hemorrhage in diffuse axonal injury, superficial siderosis, old intraventricular hemorrhage, thrombosed aneurysm, and some calcifications...

Further, You can have a look at this article, which will help you more

Article Principles, Techniques, and Applications of T2*-based MR Ima...

Thanks

Sam

Let me add some more technical details to Samson Nivins' answer:

Both T2 and T2* describe the transverse relaxation (i.e., the decay of the MRI signal induced by the precessing transverse nuclear magnetization). T2 describes the decay observed in spin-echo (or turbo-spin-echo, fast-spin-echo, RARE, HASTE, SE-EPI, ...) measurements. T2* describes the decay in gradient-echo (or FLASH, SPGR, FID-EPI, ...) measurements.

Technically, T2* includes additional (static) effects due to macroscopic and microscopic magnetic field inhomogeneities (caused e.g. by blood) and is always shorter than (or at most equal to) T2.

Mathematically, this is expressed by 1/T2* = 1/T2 + 1/T2' (or, equivalently, T2* = (T2 × T2')/(T2' + T2)), where T2' describes the transverse relaxation only due to static magnetic field inhomogeneities.

The mathematical description is already given above and is written in many MR books (for example, in "Handbook of MRI Pulse Sequences" (Authors: Bernstein, King, Zhou)).

My colleagues apply T2* mapping for cartilage imaging, as for instance

Article T2*-Mapping of Acetabular Cartilage in Patients With Femoroa...

thank you for your detailed answers.

The differece between T2 and T2* is due to the magnetic field inhomogeneities (Spin echo - Gradient echo). Combinations of the T2 and T2* along with the diffusion and perfusion MRI parameters may be used to characterize the capillary space.

One must also be aware that any decaying signal (e.g., the right side of an echo or gradient echo signal , or FID) will always be T2*.

It is only the left hand side of the echo that is T2 (i.e., the rephasing of spins, which is T2).

This also implies that the full echo signal is not symmetric.

This also explains why T2 can only be estimated by modeling the amplitude value of a series of echos (i.e., from the CPMG sequence).

Mr. Stanley, it is actually the opposite.

The spin echo rephases the T2* effects, one can not rephase the effect of T2, therefore the left side of the echo (the rising part) also shows the effect of T2* (the spins that dephased due to T2* are "getting back together", resulting in measurable transverse magnetization)

Thereby the echo is almost symmetrical (consider the popular analogue for the SE pulse, stating it is like reversing time - the left side of the first echo is identical to the initial FID in this regard - it is it's time-mirrorred pair. The reason it is not completely identical is due to T2, diffusion, or other, not time-invariant effects).

Agreed, with very short T2 (order of 1-10ms), the spin echo may appear somewhat asymmetrical, since T2-decay takes place during the echo as well and is superimposed on the T2* (consider the addition formula), but in practical cases you are not seeing this.

Consider the case of a CPMG experiment, where the maximum values of the consecutive echos are decreasing accoring to T2, while you see how short T2* actually is with the narrow shape of the echos.

There is definitely a confusion.

By definition, spin-spin T2 relaxation is dependent on events of exchange of energies between two spins (e.g., spin up and spin down exchanging to become spin down and spin up) and the time between events is the correlation time, τC, from the original Bloembergen, Purcell and Pound 1948 paper. T2* includes both T2 plus the inhomogeneity term.

Any measured decaying signal (FID or right hand side of an echo) will follow the T2* time constant, which is why one cannot estimate the T2 from an FID.

If the CPMG sequence (or echo train sequence with relatively short echo spacing to minimize the diffusion effect) is the only method to measure T2, then the signal amplitude at each echo (i.e., the signal being rephrased) must reflect the T2 relaxation because it is the amplitude at each echo that is used to model the T2 time constant.

Therefore, if T2* is not equal to T2 then the time constant of the rephasing signal (left side) must be different than the decay time constant (right side).

Does this help to clarify.

Dear Mr. Stanley,

I assume that your reply was meant for me, altough You did not bother to address me.

I agree my previous reply was misleading, since I also made a common mistake adressing T2' effects as T2*, however, since, in most cases, T2>>T2', and 1/T2* = 1/T2 + 1/T2', observed T2*-decay is governed by T2'.

Nevertheless I also clearly acknowledged the fact, that T2-effects are always present in T2*, therefore the echo is not symmetrical.

I recognize the fact, that You realize the difference between the usage of an FID and a CPMG experiment, however, since neither the original question nor my reply mentioned anything even close to 'estimating T2 from an FID', I do not see the neccessity to once again emphasize this part of Your knowledge. No arguement on this matter from my part.

Indeed, the echoes are (theoretically) not symmetrical, but for practical considerations, they might as well be, however stating that the rephasing of spins (the left hand side of an echo) is related to T2, as you did in your first answer, was completely wrong, therefore I felt the need to correct that statement. The correct way to describe this phenomenon is indeed how you did in your later answer: the peaks of amplitude of the consequtive echoes in a CPMG experiment follow the T2-decay.

Regarding the paper You referenced: Bloembergen, Purcell and Pound used the τC exchange time in relation to T1-relaxation; and since spins perpendicular to an external magnetic field cannot cause changes in energy and therefore spin-spin relaxation is not accompanied by energy exchange, I would recommend the following chain of thought:

The reason why the spin echo does not reduce the T2 effects on the measued signal, lies in the rapid time fluctuations in the intrinsic local fields. (Caused by the spins.) The inhomogeneities in these internal fields do not stay fixed in time after the π-pulse, the rates at which phase is accumulated change with time, and, in general, no refocusing is possible.

Once again: effects behind T2 are not time invariant (just as motion, diffusion, or spin-lattice interaction are not time invariant), but the effects behind T2' are, therefore the latter can be reduced by the spin echo while the former cannot.

With this, I believe these problems are indeed closer to be clarified.

Best regards,

Gyula Gyebnár