ABSTRACT
In medical contexts, size change is calculated for a second scan relative to a baseline initial measurement.
e denition for volumetric relative size change CV is given by
C
V V V
1 (11.1)
erefore, for example, if a lesion increases in volume from V1 = 100 mm3 to V2 = 150 mm3, then the relative size change is 0.5 or 50%. Further, if V2 = 50 mm3, then the relative size change is −0.5 or −50%. Given a measured lesion diameter D1 and a later diameter measure D2, the relative change in diameter CD may be dened as
C
D D D
1 (11.2)
It is important to note that some of the literature reports on sizes and change-in-size based on unidimensional measures and that the fractional size change by diameter measurements are dierent than the fractional size change by volumetric measurements. ese dierent measurements may be related to each if we assume that change is isotropic in all three spatial dimensions (i.e., the change amount is the same proportion in all directions) by the following:
C C C CD V V D≈ + − ≈ +( ) −1 1 1 13 3, (11.3)
Change measurement is a key diagnostic and disease evaluation tool in common use in medical practice; however, the quantitative treatment of change measurements is not well understood and harbors some misconceptions by the medical community. In making an estimate of change, we are interested in both the value of the measured change and the uncertainty in that measurement; without knowing the latter, the measurement itself has very limited clinical value. For example, consider that there is a measured change of 7% increase in the size of a lesion; this would normally reect a signicant biological change of progression; however, if the uncertainty in the measurement is 10%, then there may
actually be a reduction in the lesion size. Unfortunately, in the context of measuring change in the size of lesions in images, measurement errors much larger than 10% are common.
