Appeal No. 1996-3409
Application No. 08/092,543
not explained why frequency distribution probability analysis would have been selected
over other models, such as linear regression analysis, stepwise regression analysis, or
cluster analysis, which cannot successfully distinguish between disease and non-disease
populations.
In our judgment, the reason advanced by the examiner for using frequency
distribution analysis in the claimed screening method (“. . . frequency distribution would
have shown distinct classifiable differences . . . “) stems from appellant’s description in the
specification, and not from the prior art. Accordingly, the rejection of claims 1 through 20
under 35 U.S.C. § 103 as unpatentable over Matson 1987 and Seltzer is reversed.
Claims 1 through 20 also stand rejected as obvious over Miyagi, Long, Seltzer and
the admitted state of the prior art.
Miyagi discloses a method of screening for disease by comparing a two-
dimensional pattern diagram representing a test subject’s integrated values of
chromatographic peaks and retention times, with a reference data base of two-
dimensional patterns generated the same way.
Long teaches that liquid chromatography, followed by electrochemical detection
and analysis of the effluent, is conventional. At pages 2 through 5 of the specification,
11
Page: Previous 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Next
Last modified: November 3, 2007