Biomedical Engineering Reference
In-Depth Information
For example, pieces of informations missing from the original ECG interpreta-
tion protocol like it is not given that how many leads should hold particular property
during diagnosis. As per our solution, we have applied test for particular properties
in all leads. This results in a characterisation of the circumstances under which the
property holds. The obtained characterisation is analysed by the medical experts un-
der all the possible conditions, and it can be used either to redefine the property or
to improve the original ECG interpretation protocol text by documenting the cases
under which the property does (or does not) hold.
More importantly, numerous anomalies became apparent during the Event-B
modelling of the ECG interpretation protocol. Here, we have used term anomaly
to refer to any issues that are not able to represent satisfactory of the original ECG
interpretation protocol. Some set of anomalies, which have found during the devel-
opment of the system are described below. We have grouped all anomalies in three
well known general categories: ambiguity, inconsistency and incompleteness.
10.6.1 Ambiguous
Ambiguous is a well-known anomaly in the area of formal representation, and it
is very hard to interpret. For instance, a problem we encountered while modelling
the ECG interpretation protocol is determining whether the terms “ST-depression”
and “ST-elevation” had the same meaning or not. These are terms that are used in
the ECG interpretation original protocol, but not defined elsewhere. Similarly, what
is the difference between “ischemia”, “definite ischemia”, “probable ischemia” and
“likely ischemia”.
In the ECG interpretation, there are 12 leads ECG signals, which are used for
interpretation, but a lot of places in the original document not clarify in which lead
the particular property should hold. Such kinds of information are very ambiguous
and give lots of confusions to model the system.
10.6.2 Inconsistencies
Inconsistencies are other kinds of anomalies which are always given conflicting re-
sults or different decisions on same patient data. The problems derived from incon-
sistent elements are very serious and as such must be avoided during development.
The ECG interpretation protocol presents several inconsistencies. For instance, we
found an inconsistency in form of applicable conditions in the ECG protocol. It ex-
presses that the conditions are applicable to both “male” and “female” under some
certain circumstances. However, elsewhere in the protocol an action is advised that
these conditions of the protocol are not applicable to “female”.
Search WWH ::
Custom Search