Explaining Medical Model Exceptions - Computational Intelligence in Healthcare: Advanced Methodologies - page 268

Since estimated solutions are usually based on domain independent interpolation,
extrapolation, or regression methods, a medical expert is not involved. An estimated
solution is not considered as full reconstruction but just as estimation.
A binary solution is a partial reconstruction of a missing value. Sometimes ISOR is
not able to construct either an exact or an estimated solution, but the expert may draw
a conclusion about an increase or a decrease of a missing value. So, a binary solution
expresses just the assumed trend. “1” means that the missing value should have in-
creased since the last measurement, whereas “0” means that it should have decreased.
Binary solutions are used in the qualitative models of ISOR.
4.2.2 Examples
The following three typical examples demonstrate how missing data are imputed.
First example: Exact solution
The value of hematocrit (HT) after six months is missing. Hematocrit is the propor-
tion of the blood volume that consists of red blood cells. So, the hematocrit measure-
ments are expressed in percentage.
The retrieved solution (fourth column of table 3) requires two additional parame-
ters, namely plasma volume (PV) and the weight of the patient. For the query patient
these values (measured after six months) are weight = 74 kg and PV = 3,367. These
values are inserted in the formula and the result is a hematocrit value of 30%.
This imputation is domain dependent, it combines three parameters in such a spe-
cific way that it can not be applied to any other parameters. However, the formula can
of course be transformed in two other ways and so it can be applied to impute values
of PV and the weight of the patient. The formula contains specific medical knowledge
that was once given as a case solution by an expert.
Second example: Estimated solution
This is the same situation as in the first example. The value of hematocrit that should
have been measured after six months is missing. Unlike the first example, now the PV
value that is required to apply the domain dependent formula is also missing. Since no
other solution for exact calculation can be retrieved, ISOR attempts to generate an
estimated solution. Of course, estimated solutions are not as good as exact ones but
are acceptable. ISOR retrieves a domain independent formula (fourth column of table
3) that states that a missing value after six months should be calculated as the sum of
two-thirds of the value measured after three months and one-third of the value meas-
ured after twelve months. This general calculation can be used for many parameters.
Third example: Binary solution
The value of parathyroid hormone (PTH) after six months is missing and needs to be
imputed. The retrieved solution involves the initial PTH measurement and the addi-
tional parameter phosphorus (P), namely the measurement after six months, P(6), and
the initial measurement, P(0). Informally, the solution states that with an increase of
phosphorus goes along an increase of PTH too. More formally, the retrieved solution
states: