The knowledge base of the system is featured with global
disease ontology, which indicates cases and the corresponding
treatment. The medical resources ontology covers doctors,nurses, rehabilitation devices, ambulances, and other medical
resources [24]. The ontology is better structured domain knowledge.
The classes and sub-classes in the ontologies are selected
according to doctors’ suggestion with the consideration of
information integrity, nonredundancy, and legibility to ensure
that the automated process is efficient and accurate. Figs. 4 and 5
show the two ontologies we use in the rehabilitation system [29],
[30]. Fig. 4 shows the disease ontology and its hierarchy graph
based on the similar cases in the database and doctors’ suggestion.
In this way, we can take the same or similar treatment method used in the cases to take care of the new patients. The
ontology is built here to set up a standard depending on which
the comparison will be made [31]. Fig. 5 shows the resource
ontology in which different medical resources including human
resource (e.g., doctor and medical staff) and device resource
[e.g., continuous passive motion (CPM) device and rehabilitation
robot] are the lower classes. Service time and frequency are
the information needed when using the device. In this way, the
system can choose appropriate resources quickly according to
treatment requirements derived in the previous step