The funny thing about the op's question is that the MPRAGE is not a functional MRI scan at all. The MPRAGE sequence provides T1-weighting that is typically only used for structural information about the scan. It is a nice sequence in that you can run it quickly (1-5 minutes typically) and get a high resolution image of the entire brain. From this you can ask questions like: is this brain grossly normal. You might laugh but our MR tech once found a pituitary adenoma on an otherwise normal volunteer with it one time. Anyhow, the scan is used for research purposes for overlaying statstically significant pixel changes in the actual fMRI scan (as the raw fMRI images look really, really crummy, so nobody ever shows them) and sometimes for things like image coregistration or structural volumetry.
http://www3.interscience.wiley.com/journal/112146729/abstract
The sequence itself is an inversion recovery prepared fast GRE readout with low flip angle. The reason it is used is because it provides good T1 contrast of the entire brain in a fast imaging time. Some people prefer to use a normal FGRE (short TR, low flip angle, no IR) for this purpose and I've seen abstracts/papers comparing the two. I almost think the choice is personal preference, though in the paper I cited Mugler makes a strong case for the superiority of MP-RAGE at 1.5T (but who's doing fMRI at 1.5T anymore...).
The TR itself refers to the spacing between the inversion recovery pulses. Here we tend to use 1600-1800 at 1.5T and 2000-3000 at 3T to allow for T1 recovery (lower is faster but hurts SNR). Oh, and the units on this is msec.
From there there is a number of small flip angle excitations. If I recall correctly, on the Siemens implementation, per TR (IR TR) it does a low-flip angle excitation for each slice and does a single phase encode during that big TR.
TE is the time from the small flip angle excitation until the center of k-space is reached (also in msec). This is somewhat inconsequential but should be minimized to avoid T2* weighting (but this is dependent mostly on bandwidth which affects SNR).
ST is the slice thickness for each of the brain slices, probably in mm. 1x1x1mm resolution is common for MP-RAGE at 3T.
Z: No idea. It's probably an axis determination.
L: and W: are just the windowing that you're looking at as described earlier. It has nothing to do with the actual scan. Each of the pixels coming out of the scan have an arbitrary, but relative, intensity, and the L and W just determine the color (typically in greyscale) the post-processing program shows you those numbers as. I can never understand why DICOM standard is 12-bit (unsigned). 12-bits?! So the largest possible value is 4095. Helloooooooo, can we for the love of god increase this to say 24-bit so we can be in line with our ADCs?! Sorry, tangent...
I thought MP-RAGE was a GE term too though this site says otherwise:
http://www.medcyclopaedia.com/library/topics/volume_i/g/gradient_echo_pulse_sequence.aspx. In any case, IR prepared GRE is very common for this purpose and exists on all the scanners. On Siemens scanners it's often referred to as turboflash (tfl). I think I read in the pulse sequence handbook one day that MP-RAGE specifically refers to 3D while tfl can be 2D or 3D. Anyhow, one of the most confusing things about this whole game is that different vendors have different names for the EXACT SAME THINGS. The researchers are no better. You stick a new pulse on something and you don't add a letter or a word to an existing sequence, but invent an ENTIRELY NEW NAME. Everyone thinks marketing is most important for their particular sequence, so you come out with all these weird and catchy names for things. Anyhow, given the link earlier it probably was first called MP-RAGE and then fit to the manufacturer's sequence names (GRE for GE, FLASH for Siemens, Fast Field Echo for Phillips... Ugh!).
Anyhow, how should you "interpret it"? It's a T1-weighted scan. Do people use MPRAGE sequences clinically? I guess I can call myself a physicist so I have no idea.
How much more do you need to know? I can give you some book references if you need them so you can get up to speed.
