Biomedical Engineering Reference
In-Depth Information
with
MMAD
, there is no apparent approach to selection of stage groupings that opti-
mize this correlation or is even predictive of positive or negative correlation.
The slope of plots of
LPM/SPM
ratio versus
MMAD
, with
LPM
/
SPM
ratio as the
directly measured dependent variable, reflects the sensitivity of this metric towards
detecting changes in
MMAD
(the steeper the slope, the higher the sensitivity). Thus,
slight changes in
MMAD
resulted in magnified variations in
LPM
/
SPM
ratio when
the slope was steep.
Tougas et al
.
[
11
] concluded that the ratio metric
LPM
/
SPM
is superior to using
either the separate variables
LPM
,
SPM
, or grouped stages as individual metrics,
since the ratio removes the confounding influence of
AUC
of the APSD in trying to
detect changes in
MMAD
. In this context, it should be noted that
ISM
, which, as has
already been mentioned, is directly related to the
AUC
, is determined simultane-
ously as the sum of
LPM
and
SPM
.
LPM
,
SPM
, or metrics derived from grouped
stages are each influenced by both changes in
MMAD
and
AUC
. In contrast, the
LPM
/
SPM
ratio
used in conjunction with
the sum of
LPM
and
SPM
will detect sepa-
rately changes in either
MMAD
or
AUC
.
Tougas et al
.
[
11
] also verified the lack of influence of
AUC
on the
LPM
/
SPM
ratio by performing regression analysis of this ratio versus
ISM
. Table
7.3
summa-
rizes the results of these regression analyses and compares goodness-of-fit statistics
for the
LPM
/
SPM
ratios versus
ISM
to the ratios versus
MMAD
. The results, corre-
lating
LPM
/
SPM
versus
ISM
, exhibited poorer coefficients of determination (
R
2
and
RMSE/b
values). For instance, the
R
2
values for
LPM
/
SPM
versus
ISM
were 2.5-240
times smaller than those from the corresponding
LPM/SPM
versus
MMAD
correla-
tions. Likewise,
RMSE/b
values for
LPM
/
SPM
versus
ISM
were 2-3 orders of mag-
nitude larger than the corresponding
RMSE
/
b
results from the
LPM
/
SPM
versus
MMAD
correlations.
The good correlation between
LPM
/
SPM
and
MMAD
, taken together with the
absence of a correlation between this ratio and
ISM
, is further illustrated graphically
in Fig.
7.10
, by comparing representative plots for a selected OIP product,
w9kw01
(CFC suspension MDI).
7.8
Conclusions
The EDA metrics have a solid theoretical basis that has been confirmed experimen-
tally with consistent performance over a wide range of different types of OIP-
generated aerosols. They are simple to apply to CI raw data, yet have the potential
to be very sensitive to changes in the APSD in terms of both central tendency
(
MMAD
) and
AUC
.
Search WWH ::
Custom Search