Information Technology Reference
In-Depth Information
Table 25.7
Possible degrees for a “normal” disease Acute poststreptococcal Glomeru-
lonephritis (APG)
Acute poststreptococcal Glomerulonephritis (APG)
Significant hypertension
not very often
Significant edema
not very often
Significant abdominal pain
some times
Abdominal complaints
seldom
Recurrence of clinical signs
very seldom
Massive proteinuria
in the beginning
seldom
Leukocytosis
some times
Hyaline and/or
granular casts in urine
almost always
the knowledge base and what information about a patient at hand we have. In the
following section we consider how to suspect the RD under the presence of different
initial information, in our denotation
R
SD
,
R
SD
,
R
SD
or
R
t
SD
.
Under the Presence of
R
SD
,
R
SD
,
R
SD
,
R
t
SD
25.4.1
We illustrate the following description by examples. Assume, that a knowledge
base contains
R
SD
as, for example, in the Table 25.8 and corresponding numerical
form (due to the scale Table 25.3) is in the Table 25.9. In this case
Δ
=
{
d
1
,
d
2
}
,
Σ
=
{
s
1
,
s
2
,...,
s
6
}
. We assume that the RD is not in the knowledge base, it means,
that it is not in
. Notice, that technically we can synthesize the initial tables (Table
25.6, Table 25.7, Table 25.8) within one table similar to Table 25.2. Actually, we as-
sume, that a physician has already established a diagnosis (one from the
Δ
, without a
help of computer system), but has some hesitations. In this case s/he checks him/her-
self on a particular diagnosis, maybe on two, as, for example, considered in the
Table 25.8.
First assume, that the process of inference is presented in the usual way (see
(25.6)). As was discussed above (see Section 25.3.4), the results of a diagnostic
process can be classified in confirmed, excluded and possible diagnoses.
As we speak about the RD, excluded and possible diagnoses are a subject of
interest for us: we consider how “far” is this case from the common representation.
For example, if after an inference process all of the considered diagnoses from
Δ
Δ
are classified as excluded (due to our classification, done preliminary), it can be
considered as a sign for the RD for the case at hand: a patient possibly has the RD.
As we have seen from Section 25.3.4 excluded diagnoses are defined if a symp-
tom is fully present/absent with a corresponding full occurrence/absent of occur-
rence. Simply speaking, at least one element of the Table 25.9 should be 1 or 0 and
at least one membership degree of fuzzy set
S
P
should be 1 or 0 as well. It means,
