Neural computing in pharmaceutical products and process development - Computer-Aided Applications in Pharmaceutical Technology - page 156

■ release properties

(g) DTZ 50% dissolution times in the fi rst fl uid (t 50-fi rst );
(h) DTZ 50% dissolution times in the second fl uid (t 50-second ); and
(i) similarity factor (f 2 ) for hydrophilic matrix tablet.
Maps can be used to analyze the infl uence of formulation factors on
the release properties, for example, increase in the amount
of dextran sulfate leads to an increase in 50% DTZ dissolution
of the fi rst fl uid (bottom right corner) and increase in 50% DTZ
dissolution of the second fl uid is associated with lower amount
of [2-(diethylamino)ethyl] dextran (bottom left corner). The amounts
of DS and EA are positively and negatively correlated with the
binding ratio, respectively. The amount of HPMC is positively
correlated with viscosity and these factors are negatively correlated
with turbidity.

Example 2

SOMs have been used to clarify the relationships between
causal factors and response variables in development of a
transdermal delivery system for ondansetron (Obata et al.,
2010). Ondansetron hydrogels composed of hydroxyethylcellulose
and hydroxypropylcellulose as gel bases, l-menthol as a penetration
enhancer and isopropanol (IPA), -methyl-2-pyrrolidone (NMP), and
water as a solvent, were prepared. The quantities of IPA (X1),
l -menthol (X2), and NMP (X3) were selected as causal factors. An
in vitro skin permeation study and an in vivo skin irritation study
were performed with the test hydrogels, selecting the fl ux and the
total irritation score (TIS) as response variables. Optimal
formulation is the one that has an appropriate penetration (fl ux)
and acceptable skin irritation score (TIS).
Each feature map (Figure 5.17) shows the value of one variable
in each map unit. Maps of fl ux and TIS are clearly distinguishable.

￿

￿ ￿