348
G. Chuiko et al.
1 Introduction
The amount of medical data generated by Internet-of-Things (IoT) devices is
constantly growing. Transition to electronic protocols and web-based systems for
healthcare management allowed the appearance of complex tools for patient treatment, monitoring, on-line consultations, and other applications. Such systems store,
process, and devise significant loads of information. While the storage of data
is significant, the most important outcome is recommendations generated as a
result of the analysis of input information. One of the instruments to improve
recommendations is semantic technologies.
The primary goal of semantic technologies [1–4] is to provide new level of
abstraction for machine data understanding. This empowers novel techniques for
data interpretation and advanced means for recommendations. On the basic level,
inclusion of semantic technologies is marked by annotation of incoming information
and further processing using semantic engines [5, 6]. Annotated data have better
level of meaningfulness and introduce additional interface for extended analysis.
Another problem that semantic technologies are trying to overcome is interoperability [6, 10–14, 20]. IoT itself is primarily characterized by high degree of heterogeneity. Devices employ different means for information transfer, different protocols,
expose specific Application-Programming Interfaces (API) during their work, etc.
[7]. All of these facts entail the issue of interoperability between IoT-devices and
all other components. For instance, usage of two different communication protocols
in temperature sensor node demands coordination on the level of server or gateway
to observe measurements from both sensors simultaneously. Both hardware and
software layer are concerned in this case [18].
It also has direct implications for medical IoT-devices as products from various
vendors might be incompatible with a specific system, require separate application for
monitoring, etc. As has been mentioned above, semantically annotated data can serve
as an additional interface. From the point of view of interoperability, measurements
from different devices can be aggregated to form a data frame about current patient
state.
The final touching point of the presented paper is provenance [15–17] of medical
data generated by IoT-devices. It is important to note that medical data should always
be taken into account in combination with other factors e.g., previous measurements, observed patient state, patient’s health card, parameters of device itself, etc.
That creates the possibility to improve Quality-of-Service (QoS) for end users. For
instance, that could be an early identification of the disease and prevention from its
further development. Annotation of medical data with semantic information, in the
following context, is an enabler of the mentioned functions.
At the same time, semantic annotation means are typically not available on the
level of the device. They are deployed on the higher level of medical monitoring
system in forms of software services.
G. Chuiko et al.
1 Introduction
The amount of medical data generated by Internet-of-Things (IoT) devices is
constantly growing. Transition to electronic protocols and web-based systems for
healthcare management allowed the appearance of complex tools for patient treatment, monitoring, on-line consultations, and other applications. Such systems store,
process, and devise significant loads of information. While the storage of data
is significant, the most important outcome is recommendations generated as a
result of the analysis of input information. One of the instruments to improve
recommendations is semantic technologies.
The primary goal of semantic technologies [1–4] is to provide new level of
abstraction for machine data understanding. This empowers novel techniques for
data interpretation and advanced means for recommendations. On the basic level,
inclusion of semantic technologies is marked by annotation of incoming information
and further processing using semantic engines [5, 6]. Annotated data have better
level of meaningfulness and introduce additional interface for extended analysis.
Another problem that semantic technologies are trying to overcome is interoperability [6, 10–14, 20]. IoT itself is primarily characterized by high degree of heterogeneity. Devices employ different means for information transfer, different protocols,
expose specific Application-Programming Interfaces (API) during their work, etc.
[7]. All of these facts entail the issue of interoperability between IoT-devices and
all other components. For instance, usage of two different communication protocols
in temperature sensor node demands coordination on the level of server or gateway
to observe measurements from both sensors simultaneously. Both hardware and
software layer are concerned in this case [18].
It also has direct implications for medical IoT-devices as products from various
vendors might be incompatible with a specific system, require separate application for
monitoring, etc. As has been mentioned above, semantically annotated data can serve
as an additional interface. From the point of view of interoperability, measurements
from different devices can be aggregated to form a data frame about current patient
state.
The final touching point of the presented paper is provenance [15–17] of medical
data generated by IoT-devices. It is important to note that medical data should always
be taken into account in combination with other factors e.g., previous measurements, observed patient state, patient’s health card, parameters of device itself, etc.
That creates the possibility to improve Quality-of-Service (QoS) for end users. For
instance, that could be an early identification of the disease and prevention from its
further development. Annotation of medical data with semantic information, in the
following context, is an enabler of the mentioned functions.
At the same time, semantic annotation means are typically not available on the
level of the device. They are deployed on the higher level of medical monitoring
system in forms of software services.