The absolute signal intensity in an MRI image is not very meaningful. You could look at the relative signal intensity of the lesion, compared to the surrounding tissue, but that won't be very precise. The relative signal intensities depend on things line proton density, T1, T2, T2*, Tr, Te, B0, and the pulse sequence. When comparing images from different papers, it is likely that many of the parameters will be inconsistent, which makes quantitative comparisons difficult.

For quantitative comparisons parameters such as T1, T2 or T2* (depending on the pulse sequence), and proton density should be measured and compared. But measuring these parameters by traditional methods is time consuming and not often done. The recent development MR Finger Printing, has aimed to measure these parameters quickly and then generate synthetic images with whichever contrast is desired.

To add to the good answer by R. Allen Waggoner . In addition to differences in signal intensity with pulse sequence, you will also have signal changes due to different Magnets (between manufacturers and even within manufacturer), Field strength (1.5T vs 3T and others), and finally there have been shown changes in signal intensity (within scanner and pulse sequence) with age.

There may be ways to stabilize the signal measure by normalizing it on something like the SNR, or Gray/white signal ratio, but those values also vary between magnet, field, pulse sequence, and age. normalizing on some of these values will simply reduce some, but not all of the systemic variance.

Best of luck.