Biomedical Engineering Reference
In-Depth Information
out that one fruitful way to view mixed-case models is through the notion of
panel count data, which is described below.
Suppose that N = fN(t) : t 0g is a counting process with mean
function EN(t) = (t), K is an integer-valued random variable, and T =
T k;j ; j = 1;:::;k;k = 1; 2;:::, is a triangular array of potential observation
times. It is assumed that N and (K;T) are independent, that K and T
are independent, and T k;j1 T k;j for j = 1;:::;k, for every k; we inter-
pret T k;0 as 0. Let X = (N K ;T K ;K) be the observed random vector for
an individual. Here, K is the number of times that the individual was ob-
served during a study; T K;1 T K;2 ::: T K;K are the times when they
were observed and N K = fN K;j N(T K;j )g j=1 are the observed counts at
those times. The above scenario specializes easily to the mixed-case interval-
censoring model, when the counting process is N(t) = 1(S t), S being
a positive random variable with distribution function F and independent of
(T;K). To understand the issues with mixed-case interval-censoring it is best
to restrict to Case 2 interval-censoring, where K is identically 2. For this case,
I use slightly different notation, denoting T 2;1 and T 2;2 by U and V , respec-
tively. With n individuals, our (i.i.d) data can be written as f i ;U i ;V i g i=1 ,
where i = ( (1 i ; (2 i ; (3 i ) and (1 i = 1(S i U i ), (2 i = 1(U i < Si i V i ),
and (3 i = 1(V i < Si i ) 1 (1 i (2 i . Here, S i is the survival time of the
i-th individual. The likelihood function for Case 2 censoring is given by
Y
F(U i ) (1)
(F(V i ) F(U i )) (2)
(1 F(V i )) (3)
L n =
;
i
i
i
i=1
and the corresponding log-likelihood by
n
o
X
(1 i log F(U i ) + (2 i log(F(V i ) F(U i )) + (3 i log(1 F(V i ))
l n =
:
i=1
Now, let t 0 0 < t 1 < t 2 ::: < t J denote the ordered distinct observation
times. If (U;V ) has a continuous distribution, then, of course, J = 2n, but in
general this may not be the case. Now consider the rank function R on the set
 
Search WWH ::




Custom Search