Complex Measurement Systems in Medicine: from Synchronized Monotask Measuring Instruments to Cyber-Physical Systems

IT in Industry, vol. 6, no.2, 2018 Published online 17-Apr-2018
Copyright © Wasserman, Кartashev, Roudnitsky 25 ISSN (Print): 2204-0595
Zhvalevsky 2018 ISSN (Online): 2203-1731
Complex Measurement Systems in Medicine: from
Synchronized Monotask Measuring Instruments to
Cyber-Physical Systems
Е.L. Wasserman
St. Petersburg Institute for Informatics and Automation of
Russian Academy of Sciences,
St. Petersburg State University,
Herzen State Pedagogical University of Russia,
St. Petersburg, Russia
N.К. Кartashev, S.B. Roudnitsky, О.V. Zhvalevsky
St. Petersburg Institute for Informatics and Automation of
Russian Academy of Sciences
St. Petersburg, Russia
Abstract—Design problems of flexible computer systems for
physiological researches are discussed. The widespread case of
employing of commercial medical devices as parts of the resulting
computer system is analyzed. To overcome most of the arising
difficulties, we propose using of the universal synchronizing
device and the modular script-based software. The prospects of
such computer systems are outlined as an evolution of them into
cyber-physical systems with on-demand plugging in of required
hardware modules.
Keywords—cyber-physical systems; integration; aggregation;
medical systems; synchronization
I. INTEGRATION AND AGGREGATION OF MEDICAL
INFORMATION-MEASUREMENT SYSTEMS
Modern information-measurement systems (IMS) which
are used in medicine can solve a wide variety of diagnostic,
prognostic, monitoring and treatment problems. However any
highly specialized system, even being the most adapted for the
performance of a specific objective function, is limited in the
possibilities of increasing the accuracy of measurements and,
consequently, of reducing the probability of errors of the first
and second kind and making the wrong decisions by the
system and, as a result, by its user. The combining of IMS into
complexes allows to essentially expand the capabilities of
IMS, to improve their accuracy and probabilistic
characteristics and, thereby, to increase the effectiveness of
the work. This combination can be either static, not involving
a frequent modification of the complex or not allowing it at
all, or dynamic, allowing adaptation of the complex for the
changing goals.
Being combined into complexes makes IMS, as a rule,
integrated. Integration implies a reasonable elimination of
hardware and software redundancy (duplication of the same
nodes, blocks, and designs as well as software, partially). So,
for example, it becomes expedient to have a common
timekeeper and frequencies generator, some of computational
means, and sometimes a secondary power source and design
elements. Reliability, built-in control and initial installations
issues should be addressed taking into account the integrated
nature of a system. However, in the case of dynamic,
situational integration, the importance of such elimination of
hardware and software redundancy is reduced. Moreover, it
can significantly reduce the ability to change the configuration
of the complex and its reliability in the event of failure or
incorrect operation of one of its parts.
Integration should not be confused with aggregation
(complexing). In contrast to the first, at the heart of
aggregation lies the idea of beneficial use of some hardware
redundancy. Redundant devices should be measuring the same
parameters. This is especially effective in case of the
heterogeneous physical principles that underlie the functioning
of the sensors since it provides a different character of the
error spectra of the measured values. Aggregation provides
greater accuracy and robustness of estimates and greater
reliability of the whole IMS. Usually, an aggregated system
brings up the qualities that are not inherent to any of the
comprising it measuring devices.
In IMS, aggregation can be carried out at different levels,
which are different stages of registration, processing, analysis,
and interpretation of data. If in one system, while solving one
task complex processing is carried out on several levels, the
aggregation becomes hierarchical and can sometimes be
completed only at the decision-making stage. Today high-
level aggregation is usually done by software.
In general, the most effective processing is the primary
complex one (processing of directly recorded signals), since it
makes it as easy as possible to work in conditions of heavy
interferences and artifacts of different origin, which is
especially important for monitoring systems. However,
primary aggregation is not always possible (in the case, for
example, of the invasiveness or extreme expense of the
measurement procedure) and requires much more
computational cost than complex processing of a higher level.
In addition, the primary aggregation is more difficult to
implement technically and, as a consequence, the resulting
system turns out to be less flexible: a change in its design may
be required to change the set of aggregated devices. Therefore,
if it is impossible or inexpedient to implement aggregation at
the first level, and also when it is not enough to fulfill the
objective function, it is needed to organize the aggregation at a
higher level or consecutively at several levels. An example of
such aggregation (if terms “measurement” and “aggregation”
This work was supported by Department of nano- and information
technologies of Russian academy of sciences, grant No. 0072-2015-0007.

is treated as widely as possible) is one of the newest,
prospective and most difficult to implement diagnostic and
research technologies — EEG-fMRI [1, 2].
Structurally, the IMS complex can be:
A) a single device;
B) a set of devices specially adapted to co-work on the
solution of a specific task;
C) a set of specialized devices, combined in a complex
based on the synchronizer;
D) a set of independent devices having compatible
interfaces and exchange protocols for control
commands and data.
The A and B variants (Fig. 1, A and B, respectively)
require significant time and material expenses, as well as
employing qualified specialists in the field of instrumentation,
electronics and computer technology. Therefore, such
solutions are used mainly in cases where it is assumed that
serial production of highly reliable devices is expected, or if
technical conditions for some reason do not allow other
options.
Fig. 1. Variants of IMS complexes structure. (Synchronizing pulses —
dotted lines, stimuli and/or intervention — dashed lines, control actions —
thin solid lines, recorded signals — thick solid lines, asynchronous flow of
registered and preprocessed data — wide arrows)
The C variant (Fig. 1, C) gives the flexibility of the
composition of the complex at a low cost of its creation and
modification due to the possibility to use already owned
equipment, including one of different manufacturers, not
necessary readily compatible. The central, integrating element
of such a complex is a specially designed device — a
synchronizer — that provides functions that are important for
solving the task: synchronization of recorded signals,
generation of control signals, sometimes user interface, etc.
The D variant (Fig. 1, D) — a set of compatible equipment
— can include the master device, initially optimized for this
function. However, the master device can be situationally set
during the formation of the complex and the assigning of the
next task to it. At the same time, management can be both
centralized and decentralized. In the latter case, control actions
are generated by that part of the complex that (for example, in
response to detection of a given event) initiates the execution
of actions by actuators. The topologies of such complexes can
be different, depending on the distribution of functions
between devices, directions of information flows, as well as
the risk and significance of the consequences of possible
errors. Component software can also be based on different
architectures, mathematical and logical principles, but
necessarily — on a compatible protocol (or set of protocols)
for external data exchange and commands. A set of
specifications for different clinical or research tasks and
hardware compatibility are provided by the “plug-and-play”
properties of the hardware [3].
II. MEDICAL CYBER-PHYSICAL SYSTEMS
Complexes of the fourth type of situationally assembled
compliant information-measuring devices have acquired
properties of cyber-physical systems in the past two decades
(for review see, e.g., [4]). Their main difference is highly
reliable hardware and software automation, which provides
flexibility and multitasking of the entire complex without
direct user intervention. To date, medical cyber-physical
systems (MCPS) are vital, context-based, networked systems
of medical devices (for review see, e.g., [5]). These systems
are used more and more often in hospitals to provide status
monitoring and continuous automated patient care [5, 6, 7].
Moreover, the recent rapid development of inexpensive
low-power communication, sensing and impact technologies
further intensified the automation of medical diagnostics and
treatment, including the using of the MCPS. Massive
implementation of the MCPS can lead to another revolution in
medicine, giving more reliable warning systems, supporting
clinical solutions, advanced diagnostics, minimally invasive
surgical care, timely, situationally dosed administration of
drugs and physical effects on the patient's body, while
ensuring safety and high productivity [8, 9]. In a routine work
process in medicine, which relies on people who make all
decisions, the patient is often exposed to danger due to a
subjective factor — human errors and miscalculations. If the
patient's condition is stable (or varies within specified limits),
the automatic system can be used to create partially closed
circuits by controlling the equipment and/or influencing the
patient. Due to this, caregivers can focus on making the most
important decisions, even in critical situations, without being
distracted by false alarms of warning systems or on the
monitoring duties [10].

In the monitoring MCPS, measurements collected by
several devices are combined at different levels to reliably
provide early detection of critical states [8, 9]. According to
forecasts, in the next decade there will be a surge of remote
monitoring systems based on wearable monitors. These MCPS
will be able to transfer data to a private or public cloud for
storage and processing, and machine learning algorithms that
work in the cloud and process this data will provide decision
support to medical personnel [11]. And patients will not be
restricted by the borders of medical institutions [12].
However, at the same time, the requirements for the
MCPS, especially of their software, also changed. This is due
to the increase of depending on the functionality of MCPS
software, higher levels of aggregation of measurements,
increasing use of networking, distributed computing (grid
technologies) and, on the other hand, the growing need for
continuous monitoring of patients also ensuring a high degree
of autonomy, safety, and confidentiality [5, 6, 11, 13]. In
addition, the behavior of MCPS is characterized by a complex
nonlinear interaction between discrete computational and
continuous physiological processes [14]. In each case of
practical application of the MCPS, one need to make sure that
it is assembled correctly (by the user or automatically) and
works in accordance with its specification. In this instance,
some unacceptable situations, including dangerous ones, may
be unique for this complex and not inherent to other variants
of its configuration, and even more so — to its individual
components. To prevent these situations, or at least timely
detect them and alarm about them, special methods are
developed [3, 9, 15, 16].
In addition, unknown or unrecognized physiological
processes (human physiology is too complicated, non-linear
and not fully observable) can lead to short-term deviations
from the set clinical monitoring scenarios, causing a false
alarm or a malfunction in the system. To resolve such
situations adaptive methods are developed with the use of
computer simulation and learning [10].
When designing the MCPS, which widely uses network
and cloud technologies, the protection of medical data at each
level of such system and the patient's safety in the event of a
network failure with the switching off some devices from
information exchange become critical problems. Due to the
differences in hardware and communication capabilities of
each level, the single encryption scheme may not be
applicable [11]. The design and software of the closed
network MCPS should provide security when one of the data
streams or control signals ceases [7].
III. PROBLEMS OF SYNCHRONIZATION OF RECORDING
EQUIPMENT AND A VARIANT OF THEIR SOLUTION
In many tasks based on the processing of physiological
signals, synchronization of information-measuring devices is
required along with preservation of information about the
temporal relations during registration and storage of these
signals. The more rapid physiological processes are subjected
to analysis, the greater the accuracy of synchronization must
be ensured. In case of complex processing with aggregation at
low levels, the required accuracy can be provided only by
hardware. However, when aggregation is at a higher level,
including the use of MCPS, it can be difficult and extremely
costly to implement, especially when the independent IMS are
temporarily combined in the complex, and the integration
itself is programmed in quasi-real or post-production time. In
addition, for the creation of the MCPS, as described above, a
high degree of device compatibility is required, i.e. they must
be initially designed for permanent or temporary functioning
as parts of the complex. Therefore, if it is required to combine
the equipment that was not designed for this purpose avoiding
the high costs, the MCPS is not an option. Especially often
this problem arises in research teams, when the complex can
be assembled only once and for a brief time, and reliability
issues are of no importance, and also at the initial stages of
introduction of new diagnostic techniques into clinical practice
(before commercially grounded production of more
technologically sophisticated equipment is started).
In such cases, as well as for end-to-end data
synchronization, special solutions may be required. The united
time in which the information and measuring devices function
does not eliminate the problem of synchronization of data
already recorded on the medium, since significant delays can
occur in each device, especially if the signal is processed in a
quasi-real or post-production time. We came to the conclusion
that the use of a more or less universal synchronizer as an
integrating element would be the optimum from the standpoint
of the flexibility of the entire complex and the cost savings in
its creation [17, 18]. The rationale for this solution for the least
favorable initial situation is as follows.
Each of the recorded signals (electro-, mechano-, phono-
or videogram, stimulation parameter of the investigated object,
a mark of a significant event, an indicator of the operation of
the equipment, etc.) in the general case is a time series or can
be reduced to it with an acceptable accuracy. Ready-made
systems capable of synchronously registering all signals of
interest to the researcher can be easily selected from serially
produced equipment, but it only can be used for conducting
routine diagnostic procedures. The equipment originally
intended for solving research problems associated with the
recording of non-standard signal sets is virtually absent on the
market and is most often manufactured on a special order,
which requires a long time and high material costs.
Therefore, as a rule, information is collected by the
researcher with the use of the recording instruments at his
disposal. Currently most of these devices happens to be an
IMS, which includes a personal computer (PC) running an
operating system (OS) with time-sharing (for example,
Microsoft Windows, UNIX-based systems). In the OS
environment the application software created by the device
developers interacts with its instrumental part: it sets up the
recording parameters, receives the data stream (measurement
results), etc. This software is usually closed to the user.
All this leads to the problem of synchronization of signals
recorded by various devices, especially if the synchronization
accuracy must be high. All signals are recorded by
unconnected devices, each of which functions in its own time.
Even if you can install the software of all the devices to one
PC, or synchronize the system timers of all used PC, the time
for all the devices will not be the same as they work under the
OS with time-sharing, while they themselves are closed

systems with an unknown and, moreover, variable delay.
Thus, it is impossible to note in the recordings the time of
significant events for the study (e.g. planned external impact
— audio, visual or other stimulus, drug administration, etc.)
only by means of involved in the process devices.
In such a situation it is advisable not to change completely
the instrument base when the need in new channels appears,
but only to fill it with lacking instruments, while maintaining,
if necessary, the already available ones. However, in order to
produce multi-channel recording with the help of diverse
devices in the united time, it is necessary to synchronize all
interconnected devices. In cases when devices that are being
synchronized do not require any special measures to ensure
the compatibility on the physical layer, we propose to
implement this using a special external device — the
synchronizer. (In some cases however physical compatibility
is essential, e.g. when combining EEG equipment and
functional magnetic resonance imaging, which should be
partly spread in time and space [2, 19].) The main function of
this device is the formation of pulses for supply to the spare
inputs of the recording devices or, if no such inputs are
available, superimposing pulses on the signal measured by
devices. The proposed solution overcomes the limitations
inherent to all implementations of software synchronization,
because the pulses generated by the synchronizer are
connected to the same time scale as the signals detected by
devices; the pulses being subjected to the same initial
processing, filtering, digitization, and other transformations
that the primary measured signal is. Moreover, pulses going
through the whole signal channel allow to track the state of the
recording devices.
This synchronizer, as an open architecture device, becomes
the basis of a complex of recording devices, easily adaptable
for a variety of research tasks. In this case, specific
requirements for synchronized hardware are minimal: either
(better) to have one free digital or analog channel, or the
technical capability to overlay pulses on one of the recorded
signals. The power, duration, shape, and repetition rate of the
pulses are set based on the characteristics of devices to be
synchronized, research schemes and planned methods of
subsequent data processing.
To remove the limit of the number of devices to be
synchronized, the support for cascading at the hardware level
in the synchronizer circuit must be provided. Due to this N
synchronizers interconnected through a separate interface will
be visible from the point of view of the high-level software
and the end user as a single device having N times more inputs
and outputs than one synchronizer [17, 18].
In 2010–2012 the authors have developed and successfully
tested the device “Polygraph-synchronizer LBMI-001” [17],
which combines the functions of the synchronization of
external devices and a number of features common to
traditional polygraphs as recorders of physiological signals.
To be able to change and expand the list of devices to be
synchronized, several free channels with customizable features
to a set of inputs and outputs of the synchronizer were added.
Scope of IMS on the basis of such synchronizer is broader
than purely scientific research: firstly, if diagnostic task being
solved in a medical facility has not been established as a
routine procedure yet, considerable material and technical
means can be saved by recombination of equipment while
maintaining the core of the complex and some plugged in
devices (modularization principle); secondly, after the
effective (from the point of view of diagnostics) configuration
options are found and tested, the IMS can be commercialized
by releasing this successful configuration as an independent
diagnostic device with a predetermined set of functions.
IV. SOFTWARE OF FLEXIBLE COMPLEXES BASED ON THE
INTEGRATOR-SYNCHRONIZER
The functions of considered flexible situationally
(re)configurable systems do not include only the collection of
“raw” data. Data processing in the course of research is mainly
carried out not in real time, because the methods capable of
solving a specific problem are not always known beforehand.
Nevertheless, some computations should be performed
immediately at the time of measurement. Such computations
include, for example, the estimation of traditional medical
diagnostics average characteristics of the patient, calculated
using a sliding time window (heart rate, respiratory
movements, etc.), recognition of pathological patterns,
requiring intervention, in monitored signals or parameters of
the signals on which depends the further experiment course. It
is not always possible to determine the desired set of
computational procedures in advance, so they should be
embeddable in the software of the complex during its usage.
Thus, the architecture of the software should be modular, with
the possibility of adding when necessary the existing new
modules, as well as specially developed ones.
However, the requirements for the software are not limited
by modularity: while using the same configuration of
equipment, different research scenarios can be implemented.
When software is developed it is difficult or impossible to
determine what sequence of functional tests, which set of
parameters measured, and what conditions an operator-
physiologist will choose. The wider range of complex
functions, the greater the uncertainty. Due to the extensibility
and cascading of “LBMI-001”, a complex built on its base
demonstrates one of the extreme variants of this uncertainty.
On the other hand, the wider the flexibility of architecture of
the complex, the higher the probability of incorrect
configuration of complex and of appearance of technical
artifacts in the recorded data. Respectively, the higher level of
expertise required from its user, too.
Uncertainties described (hardware configuration, scenario
of studies and methods of operative data processing) lead to a
number of problems in the development of software; the most
significant include the choice of the user interface and the
choice of storage structures for data obtained during the
investigation.
When designing the user interface it is needed to be aware
that the IMS end user, as a rule, is not an expert in information
technology. Therefore, the most obvious solution for
providing interface flexibility — integration of programming
language into the software and providing an access to the
basic functions of the complex and graphic users controls —
in this case is not sufficient. It is desirable to use a higher level
of abstraction and offer the user a kind of “interface

constructor”. A higher level of abstraction is the script that
controls the underlying storage system and user interface.
The storage structure for data obtained in investigations is
also requires versatility. It requires from IT specialist, serving
the complex, a good knowledge of the characteristics of data
generation in each software module, otherwise even the
slightest mistake can go unnoticed and give rise to secondary
artifacts at the stage of data processing and analysis. A better
solution is to develop a universal storage structure that
accommodates all possible scenarios of the use of the
complex. At the same time, giving the complexity and
multidimensionality of (and sometimes — the uncertainty and
omissions in) the data, unique identifiers for the patient, the
patient's visit, test in the framework of the visit, a set of the
similar data in the test are necessary.
In addition, to minimize the chance of missing a
significant event when using such a flexible complex, it is
necessary to provide for recording of the two protocols:
unswitchable off complete log of all events (including the
actions of the user and the configuration and condition of the
equipment), as well as customized user protocol, recorded at
his request.
To improve the quality of the physiological measurements,
in many cases it is desirable to provide a sufficient level of
automation, which minimizes the influence of random factors
on the characteristics of presented stimuli, feedback control,
and the measurement processes. Consequently, every new test
scenario must be supported by the software, preferably at the
level of the user interface. For example, if, according to the
experiment plan, various functional tests are performed
sequentially, the automation of firing of respective software
modules will not only provide equal delays between tests in
different patients, but also eliminate the human factor. When,
for example, new types of physiological signals appear, or the
known tests should be repeated, the system continues to
operate transparently for the user, when the underlying
systems are restructured automatically.
There are at least two types of script: recording and
processing ones. Wherein processing scenarios are divided
into operative — implemented in the real-time — and delayed:
in the former processing and analysis of the known methods is
performed at once (e.g., to provide information to the patient
or for archiving), and in the latter some part of the recorded
data is set aside for further processing and in-depth analysis.
V. CONCLUSION
Although so far it is not possible to create a fully universal
system and research software using closed parts produced by
different manufacturers, nor to reach the technological level of
medical cyber-physical systems in full, the proposed approach
to the construction of the flexible architecture of research
complexes and the introduced levels of abstraction allow to
overcome the main difficulties. It provides the researcher with
a fundamentally new opportunity — easy handling of all the
available instruments as an adaptive (to the current task)
matrix of diagnostic resources. The key features of such
information-measurement system must be:
● integration of the recording, control and actuating
apparatus on the basis of an external synchronizing
device;
● cascading of synchronizers supported at the hardware
level;
● use of configurable pulses, that propagate along the
same channel the recorded signal does;
● a modular software architecture with a uniform
interface (adapted to the tasks of each module) and the
ability to plug in the new modules;
● use of custom user scripts of the investigation
involving different hardware configurations and
different algorithms for data collection and
processing;
● universalization of identifiers system and databases of
recorded data.
REFERENCES
[1] J. Gotman and F. Pittau, “Combining EEG and fMRI in the study of
epileptic discharges,” Epilepsia, vol. 52, suppl. 4, pp. 38-42, Jul. 2011.
DOI: 10.1111/j.1528-1167.2011.03151.x.
[2] S.F. Storti, E. Formaggio, E. Franchini, L.G. Bongiovanni, R. Cerini,
A. Fiaschi, C.M. Michel, and P. Manganotti, “A multimodal imaging
approach to the evaluation of post-traumatic epilepsy,” MAGMA, vol.
25(5), pp. 345-360. Oct. 2012. DOI: 10.1007/s10334-012-0316-9
[3] K.K. Venkatasubramanian, E.Y. Vasserman, V. Sfyrla, O. Sokolsky, and
I. Lee, “Requirement engineering for functional alarm system for
interoperable medical devices”, In Proceedings of the 34th International
Conference on Computer Safety, Reliability and Security (SAFECOMP
2015), The Delft, The Netherlands, pp. 252-266, Sept. 2015. DOI:
10.1007/978-3-319-24255-2_19
[4] R. Rajkumar, D. Niz, and M. Klein, Cyber-Physical Systems. Boston,
MA: Addison-Wesley, 2016.
[5] I. Lee, A. Ayoub, S. Chen, B. Kim, A. King, A. Roederer, and
O. Sokolsky, “Medical cyber-physical systems,” in Cyber-Physical
Systems, R. Rajkumar, D. Niz, and M. Klein, Boston, MA : Addison-
Wesley, 2016, pp. 3–60.
[6] I. Lee, O. Sokolsky, S. Chen, J. Hatcliff, E. Jee, B.G. Kim, A. King,
M. Mullen-Fortino, S. Park, A. Roederer, and K. Venkatasubramanian,
“Challenges and research directions in medical cyber-physical systems,”
Proceedings of the IEEE, vol. 100(1), pp. 75-90, Jan. 2012. DOI:
10.1109/JPROC.2011.2165270
[7] M. Pajic, R. Mangharam, O. Sokolsky, D. Arney, J.M. Goldman, and
I. Lee, “Model-driven safety analysis of closed-loop medical systems,”
IEEE Transactions on Industrial Informatics, vol. 10(1), pp. 3-16, Oct.
2012. DOI: 10.1109/TII.2012.2226594
[8] J. Weimer, R. Ivanov, A. Roederer, S. Chen, and I. Lee, “Parameter-
invariant design of medical alarms,” IEEE Design and Test, vol. 32(5),
pp. 9-16, Oct. 2015. DOI: 10.1109/MDAT.2015.2451083
[9] J. Weimer, O. Sokolsky, and I. Lee, “Robust medical monitor design,”
International Conference on Healthcare Informatics (ICHI), Dallas, TX,
pp. 445-445, Oct. 2015. DOI: 10.1109/ICHI.2015.63
[10] S. Chen, O. Sokolsky, J. Weimer, and I. Lee, “Data-driven adaptive
safety monitoring using virtual subjects in medical cyber-physical
systems: a glucose control case study,” Journal of Computer Science and
Engineering, vol. 10(3), pp. 75-84, sept. 2016. DOI:
10.5626/JCSE.2016.10.3.75
[11] O. Kocabas, T. Soyata, and M.K. Aktas, “Emerging security
mechanisms for medical cyber physical systems,” IEEE/ACM
Transactions on Computational Biology and Bioinformatics, vol. 13(3),
pp. 401-416, May-Jun. 2016. DOI: 10.1109/TCBB.2016.2520933
[12] J. Jezewski, A. Pawlak, J. Wróbel, K. Horoba, and P. Penkala, “Towards
a medical cyber-physical system for home telecare of high-risk
pregnancy,” IFAC-PapersOnLine, vol. 48(4), pp. 466-473, Dec. 2015.
DOI: 10.1016/j.ifacol.2015.07.079

[13] I. Lee, and O. Sokolsky, “Medical Cyber Physical Systems,” 47th
Design Automation Conference (DAC '10), pp. 743-748, June 2010.
DOI: 10.1145/1837274.1837463
[14] E. Bartocci, S. Gao, and S.A. Smolka, “Medical cyber-physical
systems,” in Leveraging Applications of Formal Methods, Verification
and Validation. Specialized Techniques and Applications. ISoLA 2014.
Lecture Notes in Computer Science, T. Margaria and B. Steffen, Eds.
Berlin, Heidelberg: Springer, vol. 8803, pp. 353-355, Oct. 2014. DOI:
10.1007/978-3-662-45231-8_25
[15] L.C. Silva, M. Perkusich, H.O. Almeida, A. Perkusich, M.A. Lima, and
K.C. Gorgônio, “A baseline patient model to support testing of medical
cyber-physical systems,” Studies in Health Technology and Informatics,
vol. 216, pp. 549-553, Aug. 2015. DOI:10.3233/978-1-61499-564-7-549
[16] L.C. Silva, H.O. Almeida, A. Perkusich, and M. Perkusich, “A model-
based approach to support validation of medical cyber-physical
systems,” Sensors, vol. 15(11), pp. 27625–27670, Oct. 2015. DOI:
10.3390/s151127625
[17] S.B. Roudnitsky, E.L. Wasserman, N.K. Kartashev, and
O.V. Zhvalevsky, “Measurement integration in physiological studies:
computer system based on external synchronizer,” (in Russian),
Biotekhnosfera, No. 3-4. pp. 72–77, 2012.
[18] E.L. Wasserman, N.K. Kartashev, O.V. Zhvalevsky, and
S.B. Roudnitsky, “Flexible architecture of hardware-software complexes
for physiological researches,” (in Russian), Izvestiya Vysshikh
Uchebnykh Zavedenii. Priborostroenie, vol. 59(11), pp. 952-958, Nov.
2016. DOI:10.17586/0021-3454-2016-59-11-952-958952
[19] J. Gotman, E. Kobayashi, A.P. Bagshaw, C.G. Bénar, and F. Dubeau,
“Combining EEG and fMRI: a multimodal tool for epilepsy research,”
Journal of Magnetic Resonance Imaging, vol. 23(6), pp. 906-920, Jun.
2006. DOI: 10.1002/jmri.20577

Complex Measurement Systems in Medicine: from Synchronized Monotask Measuring Instruments to Cyber-Physical Systems

Related slideshows

Recommended for you

More Related Content

What's hot

What's hot (18)

Similar to Complex Measurement Systems in Medicine: from Synchronized Monotask Measuring Instruments to Cyber-Physical Systems

Similar to Complex Measurement Systems in Medicine: from Synchronized Monotask Measuring Instruments to Cyber-Physical Systems (20)

More from ITIIIndustries

More from ITIIIndustries (20)

Recently uploaded

Recently uploaded (20)

Complex Measurement Systems in Medicine: from Synchronized Monotask Measuring Instruments to Cyber-Physical Systems