Realizacija servisa pametnog zdravstva i njihova integracija u koncept pametnih gradova
Author
Avdić, AldinaMentor
Janković, DraganCommittee members
Stoimenov, LeonidBogdanović, Miloš
Marovac, Ulfeta
Gajić, Dušan
Metadata
Show full item recordAbstract
The development of information technologies has contributed
to the emergence of the concept of a smart city, and one of the
components of this concept is smart health. This component is
important because smart cities directly or indirectly affect the health
of residents.
The main contributions of this doctoral dissertation are a
proposal of a software platform that integrates existing public health
services as well as new e-health services, a proposal for a uniform
way of integrating heterogeneous smart health services into a smart
city, and a proposal and implementation of methods for labelling
terms in medical texts, based on artificial intelligence, without which
part of the smart health service could not be realized in an efficient
way.
The above-mentioned software platform combines the smart
health services that will be proposed (services for information and
prevention of diseases, epidemic control, search of medical
documents, automatic questionnaire processing, a...ir pollution
monitoring, labelling of medical texts, organization of screening
programs, etc.). These services are created based on input data of
different types (sensor data on locations, air pollution, text from
documents and medical information systems and data collected by
crowdsourcing, data from relational and non-relational databases), so
it is necessary to integrate these heterogeneous services in uniform
way.
Part of the proposed e-health services is based on data
processing in medical information systems as well as medical text
documents in the Serbian language. Methods for normalization,
labelling and classification of terms in the Serbian language are a
prerequisite for the successful implementation of these services, and
within this dissertation methods are proposed and implemented whose
F1-score is 0.9082, which is an excellent result compared to methods
for this purpose in other languages. For their implementation, it is
necessary to use artificial intelligence methods, such as natural
language processing, data and text mining, machine learning, etc.
Some of the proposed e-health services are practically implemented
and integrated into the smart city concept.