Biomedical Engineering Reference
In-Depth Information
Table 8.7 Key to comparisons in Table 8.8
R 2 value used to establish limits
(a) Type I region, lower limit
(b) Type I region, upper limit
(c) Type II region, lower limit
(d) Type II region, upper limit
Table 8.8 Qualitative comparison of OCCs for OIP APSDs from the IPAC-RS database in the
critical regions related to type I and II errors: EDA = decision based on ratio metric ( LPM / SPM );
Gr24: decision based on stage groups 2 and 4 (excluding group 3)
HFA suspension MDI ( w9k201 )
HFA suspension MDI ( w9j901 )
R 2 = 0.9
R 2 = 0.8
R 2 = 0.9
R 2 = 0.8
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
R 2 = 0.7
R 2 = 0.6
R 2 = 0.7
R 2 = 0.6
EDA
EDA
EDA
Gr24
EDA
EDA
EDA
EDA
Gr24
EDA
HFA solution MDI ( w9j801 )
Dry powder inhaler ( w9jk01 )
R 2 = 0.9
R 2 = 0.8
R 2 = 0.9
R 2 = 0.8
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
EDA
R 2 = 0.7
R 2 = 0.6
R 2 = 0.7
R 2 = 0.6
Gr24
EDA
Gr24
EDA
EDA
EDA
Gr24
EDA
Gr24
EDA
EDA
Gr24
EDA
Dry powder inhaler ( w9k901 )
CFC suspension MDI ( w9j601 )
R 2 = 0.9
R 2 = 0.8
R 2 = 0.9
R 2 = 0.8
EDA
EDA
EDA
EDA
EDA
EDA
Gr24
EDA
EDA
EDA
EDA
EDA
Gr24
EDA
R 2 = 0.7
R 2 = 0.6
R 2 = 0.7
R 2 = 0.6
EDA
EDA
EDA
EDA
Gr24
EDA
Gr24
EDA
EDA
EDA
Gr24
EDA
Gr24
EDA
CFC suspension MDI ( w9k001 )
CFC suspension MDI ( w9kw01 )
R 2 = 0.9
R 2 = 0.8
R 2 = 0.9
R 2 = 0.8
EDA
EDA
EDA
Gr24
EDA
EDA
Gr24
EDA
Gr24
EDA
Gr24
EDA
R 2 = 0.7
R 2 = 0.6
R 2 = 0.7
R 2 = 0.6
Gr24
EDA
Gr24
EDA
Gr24
EDA
Gr24
EDA
Gr24
EDA
Gr24
EDA
Gr24
EDA
Gr24
EDA
with the overall curve closer to the ideal OCC then to either of the stage grouping-
based alternatives.
Tables 8.7 and 8.8 summarize qualitatively a comparison of these OCCs in the
critical regions related to type I (false acceptance) and type II (false rejection) errors
(Fig. 8.38 ).
The entries in the cells in Tables 8.7 and 8.8 indicate superior performance for
the indicated metrics in the critical region of the OCC. Where no significant differ-
ence is observed, a character “≈” appears. Bolded values indicate that only the indi-
cated metric is capable and that there is a serious performance issue with the
alternative metric.
 
Search WWH ::




Custom Search