< previous page page_165 next page >

Page 165
Placing parameters obtained from nonlinear regression analysis of the plasma profile (Eqn. 1) into the fractional system response profile (Eqn. 2) transforms the data into a data set similar to the fraction released from dissolution profile.
STEP 3: Compute the fractional response-time profile.
Simply making the independent variable the dependent variable and vice versa and adding limits to the now implicit variable T allows for the solution of T at various F, thus:
INDVARS: F
DEPVARS: T
PARAMS: I KA K
F = I×K×/AUC×KA/(KA-K)×((1-EXP(-K×T)/K)-(1-EXP(-KA×T)/KA))
O < T < TMAX (3)
where TMAX is the maximum time of that profile.
By simulating the data set with the parameters obtained from Eqn. 1, Eqn. 2 will yield as many F versus T pairs for the profiles as needed. In general, 10 pairs are sufficient for most straight lines.
2. Treatment of Dissolution-Time Data
STEP 4: Curve-fit the dissolution-time data to the equation in MINSQ format
INDVARS: TIME
DEPVARS: F
PARAMS: A K
T = (SQRT(TIME)-SQRT(A))×UNIT(SQRT(TIME)-SQRT(A))
F = (3×K×T-3×SQR(K×T)+(K×T)^3) (3)
where A is the lag time and K is the Higuchi release-rate constant for single face planar release.
STEP 5: Compute the fractional response-time profile.
Again, making the independent variable the dependent variable and vice versa and adding limits to the now implicit variable TIME allows for the solution of TIME at various F, thus:
INDVARS: F
DEPVARS: TIME
PARAMS: A K
T = (SQRT(TIME)-SQRT(A))×UNIT(SQRT(TIME)-SQRT(A))
F = (3×K×T-3×SQR(K×T)+(K×T)^3)
O < TIME < TIMEMAX (5)

 
< previous page page_165 next page >