вход по аккаунту



код для вставкиСкачать
i-com 2016; 15(3): 227–247
Research Article
Jasmin Dell’Anna*, Armin Janß, Hans Clusmann, Klaus Radermacher
A Configurable Footswitch Unit for the Open
Networked Neurosurgical OR – Development,
Evaluation and Future Perspectives
DOI 10.1515/icom-2016-0031
Abstract: Footswitches are used in the neurosurgical
operating room for human-device-communication every
day. However, problems, such as shifting or confusion of
footswitches, often occur due to the parallel usage of up
to 5 device-specific footswitches, resulting in a significant
burden for the surgeon. There are no footswitches available
which offer an optional central activation of different devices
from various manufacturers and a reconfiguration during
usage. Therefore, a new concept of a configurable central
footswitch unit has been developed for optional activation
of different devices in an open networked neurosurgical OR
setting. In a user-centered evaluation 9 surgeons used both,
the configurable central footswitch unit and 4 device-specific
footswitches, for a cross-over experiment in an experimental
OR setting. It shows that all surgeons were able to handle
the configurable footswitch autonomously and that efficiency in surgeon-device-communication can be increased.
Keywords: Configurable Footswitch, Neurosurgery, Human
Machine Interaction, Handling Concepts, Integrated OR,
Open Standards
1 Introduction
“Imagine in your car the brake and gas pedal would not be
where you expect it to be […].” A scenario like this would
*Corresponding author: Jasmin Dell’Anna, RWTH Aachen University,
Chair of Medical Engineering, Aachen, Germany,
Armin Janß, RWTH Aachen University, Chair of Medical Engineering,
Aachen, Germany, e-mail:
Hans Clusmann, RWTH Aachen University, Department of
Neurosurgery, Aachen, Germany, e-mail:
Klaus Radermacher, RWTH Aachen University, Chair of Medical
Engineering, Aachen, Germany,
hardly find acceptance by car drivers and certainly not by
automotive risk managers or usability experts, but this
statement, given by an experienced neurosurgeon after
21 years of work, quite well points out the daily working
situation of many surgeons.
Footswitches are components of numerous surgical
devices and, as such, integral elements of operating room theatres. They are used to release functions of electrical devices,
e. g. the drilling, the electro-cauterization device or the X-ray
C-arm. In general, the use of footswitches facilitates the work,
since they enable to use both hands exclusively for the manipulation on the patient in the surgical field, while the feet can
be used as additional input resource. But due to the technological progress the complexity of surgical interventions has
been increasing over the last decades, especially in the field
of neurosurgery, and with it the number of medical devices in
the OR. Many of these devices use a particular footswitch, and
for some surgical interventions the surgeons have to use up
to 5 or more different footswitches, which have to find space
under the OR table.
But by reason of sterility the patient has to be covered
by sterile drapes which often hinder the view on the
footswitches (Figure 1 and Figure 4). Surgeons have to
find and release them blindly, which inevitably leads to
handling errors, especially when footswitches are shifted
or after they fell off the footboard [4, 7]. The fact that
footswitches often vary in number and design of their
elements additionally impedes safe handling and causes
stress to the surgeons [16, 21].
There are a few footswitches available which enable the
control of different devices. However, these footswitches are
proprietary solutions: they only work with selected devices
of the same manufacturer and do not communicate with
devices from other manufacturers. Thus, their use is considerably limited, and this is why those footswitches are hardly
found in the OR.
In order to overcome usability limitations due to proprietary systems, open standards for device interconnection have been developed in the German flagship project
OR.NET. The device interconnection with open standards
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
228 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
Figure 1: Use of different conventional device specific footswitches
in the neurosurgical OR.
and the associated availability of data and functions
within a common network offers new ways in human-machine-interaction (HMI) in the OR. For the first time it
appears to be feasible to control different devices from
various manufacturers with only one central footswitch
unit. As a consequence of the existing problems related
to footswitch handling a prototype of a configurable
footswitch has been developed in close cooperation with
technical and clinical project partners. This paper presents
the overall development process, starting with an analysis
of problems related to footswitches in the OR and special
requirements resulting from the neurosurgical context of
use in chapter 2. The state of the art of footswitches and
integrated operating room systems is presented in chapter
3, followed by a comprehensive requirement analysis in
chapter 4. The development process of the prototype,
based on DIN EN ISO 9241-210, is described in chapter 5
and the user-centered evaluation of the prototype is presented in chapter 6. In chapters 7 and 8 the results are discussed and future perspectives are given.
2 Background and Motivation
This section presents the relevant medical background
and problems related to footswitch handling in the OR.
Subchapter 2.1 describes the working field of Neurosurgery in general and a posterior cervical decompression
and fusion operation in more detail, which will be the
medical application context for the usability evaluation in
chapter 6. Subchapter 2.2 gives an insight into footswitch
handling in the OR and related problems, based on observations and an online survey.
2.1 Medical Background
Neurosurgery covers diagnosis, conservative and surgical treatment of diseases and malformations or injuries of
the central and peripheral nervous system. The treatment spectrum covers acute injury, tumor, infection,
and malformation of the scull, the brain and the spinal
cord. This includes cerebral or spinal hemorrhage and
vascular diseases, but also the whole range of problems
related to vertebral discs, vertebral bodies, and spinal
malformations [15]. A major challenge in the field of
neurosurgery is that manipulations always possibly
affect very sensitive tissue, and treatment mistakes can
easily cause essential damage to the patient, such as
paralysis, loss of memory, and cognitive function, or
even death.
Neurosurgeons can resort to a large amount of
medical devices in order to achieve this broad range of
treatment. The standard equipment of a neurosurgical OR
is listed in table 1, which can be supplemented by more
specific devices. For complex interventions, such as a
spinal decompression and fusion operation, up to seven
of the named devices are in use at different steps during
the operational workflow, of which five devices are controlled by use of their particular footswitches.
The development of a first prototype of a configurable footswitch unit for neurosurgical applications
should consider the device usage for the whole bandwidth of neurosurgical applications. However, for a
usability evaluation of the prototype with focus on the
medical application context it is essential to choose
a specific intervention in order to simulate a realistic
workflow. The intervention chosen is a posterior cervical decompression and fusion operation, which is
an established procedure in neurosurgery and a representative example for a demanding and high-risk
Cervical decompression of the spinal cord is necessary, if degeneration or deformities of the ligamentous
and / or osseous parts of the cervical spine lead to a
constriction of the spinal canal (Figure 2). Compression
of the cervical spinal cord results in typical symptoms
with neurological deficits like ataxia, crippling and
paralysis of the extremities [5]. For a dorsal decompression of the spinal cord the respective ligamentous and
osseous parts of a vertebra (spinous process, laminae,
medial parts of facet-joints, and interlaminar ligaments)
are removed and the spinal dura covering the cord is
exposed and decompressed. The extensive removal of
material in multiple levels may lead to a progressive
instability of the cervical spine. For this reason several
vertebrae are interconnected with titanium screws and
rods for a firm fixation (osteosynthesis), and osseous
material may be attached to this construct to enable
bony fusion.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR 229
Table 1: Standard equipment of a neurosurgical OR [15].
Medical device
Input device
Endoscopic system
Minimal invasive diagnosis and treatment
High frequency (HF) surgical device
Coagulation and cutting of soft tissue
footswitch / handswitch
High speed drill
Drilling or milling of bone tissue
footswitch / handswitch
Intraoperative neuromonitoring device
Intraoperative monitoring of neurophysiological nerve activity
mouse /key board /
touch screen
Laser system
Removal of tissue, e. g. tumorous tissue
Neuronavigation system
Computer assisted planning of the intervention, and
intraoperative transfer of the planning to the patient.
E. g. for tumor segmentation and removal
mouse / key board /
touch screen / footswitch
OR microscope
Magnification of the operative site
Handswitch / mouth-switch
(sometimes additional
footswitch available)
OR table
Positioning of the patient
remote control
Ultrasound ablation device
Cutting of soft tissue
footswitch / handswitch
Ultrasound diagnosis system
E. g. for localization of tumors or accumulations of blood
Footswitch / handswitch
X-ray device (C-arm)
Pre- and intraoperative imaging
Footswitch / handswitch
Figure 2: Left: Narrowing of the spinal canal by osseous and ligamentous structures resulting in compression of the spinal cord.
Right: Left-right and anterior-posterior X-ray images of implanted lateral mass screws and titan bars.
Figure 3: Selected steps of a cervical decompression and fusion operation (dorsal view): a) cervical spine is exposed, spinous processes are
removed, lateral mass screws are introduced, periostal bone is being removed; b) the spinal dura is exposed, heads of lateral mass screws
are aligned and titan bars are fixed, osseous material is attached laterally to the small facet joints; c) a redon drainage is laid and the wound
is being closed.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
230 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
2.2 F ootswitches in the OR – Inconvenient
Yet Indispensable
Every neurosurgeon uses footswitches during surgical
interventions. There are at least 6 different devices found
in a neurosurgical OR, which can or must be controlled by
use of a footswitch (Table 1). The HF device for monopolar cutting and bipolar coagulation is probably the most
frequently used device in neurosurgical interventions. It
is generally released by a two pedal footswitch with the
distinctive color scheme of yellow and blue, but does
often offer an additional handswitch on the instrument
for monopolar cutting. The same holds for some milling
and drilling devices, which can be controlled by both,
footswitch and handswitch. The fluoroscopic X-ray device
can be released by both, footswitch and handswitch. The
operation microscope is most often operated by handswitches which are situated directly on the device, and
offers a footswitch as an additional handling alternative.
According to a nationwide online survey, which has been
done in July 2014 with 34 neurosurgeons, problems related
to the usage of footswitches occur rather often (20 %) or very
often (12 %), while none of the 34 surgeons stated that problems never happen. The most frequently occurring problem
(with 1 = never until 5 = very often) is that a footswitch cannot
be reached by the surgeon (mean 3.5), followed by shifting
of the switch during usage (mean 2.8), the activation of the
wrong device by confusion of the footswitches (mean 2.7),
the activation of the wrong pedal on the right footswitch
(mean 2.6), tilting of a footswitch (mean 2.6), falling down
off a footboard (mean 2.5) (Figure 4 a), tripping over cables
(mean 2.4) and the accidental activation of a second pedal
(mean 2.2). A workflow analysis of 25 interventions, done in
the neurosurgical OR at the University Hospital in Aachen
in 2013, affirms this assessment: 27 problems related to
footswitch handling have been observed, and shifting or
falling off were the main causes [4]. Van Veelen et al. [21]
and Matern et al. [16] made similar observations.
The consequences resulting from each of the listed
problems had to be rated by the surgeons in the online
survey with 1 = unproblematic, 2 = marginal, 3 = critical
and 4 = catastrophic. It showed that the opinions strongly
diverged, but two of the named problems were generally
seen more critical than the others. 12 of 34 surgeons did
expect critical consequences for an accidental release of
a second device by simultaneous activation of two pedals.
This problem could also be observed by the author of this
article, when a footswitch fell off the footboard and x-ray
images were taken by accident (and unnoticed) each time
the HF device was used.
The surgeons also had to state how disturbing the
above mentioned problems are in their daily work. The
most disturbing problem is that a footswitch cannot be
reached (n = 26), but shifting and falling down from a
footboard was also named by more than 20 surgeons. The
other mentioned problems are also rated to be rather or
very disturbing by more than 50 % of the surgeons. Need
for improvement of the situation is seen by the majority of
surgeons, where 14 surgeons see high or very high need
(Figure 5).
3 State of the Art
This chapter presents the relevant state of the art for the
development of a configurable footswitch in an open
networked OR. In subchapter 3.1 available Integrated
Operating Room Systems and related research projects
are presented, being the prerequisite for the usage of a
configurable footswitch. Subchapter 3.2 addresses foot
control. It gives a description of footswitch elements and
Figure 4: Problems with footswitches in the OR: a) The footswitch fell off the footboard; b) and c) Footswitches are hidden behind sterile
drapes and not visible to the surgeon.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR 231
Figure 5: Footswitch situation at present: need for improvement.
their characteristics and presents some approaches for the
control of several devices by one footswitch.
3.1 IORS – Integrated Operating Room Systems
In integrated operating room systems (iORS) all devices
are integrated in a common network which enables data
exchange and device control. Central user interfaces offer
the possibility to check and change device parameters
without direct interaction with the device itself, and the
connection to the clinical information systems enhances
the workflow management [18]. There are several iORS
commercially available, for example the OR1 (Storz
Medical AG), the CORE nova system (Richard Wolf GmbH),
the BrainSUITE (Brainlab AG), the iSUITE (Stryker) or
ENDOALPHA (Olympus Europe SE & CO. KG). All these
systems are approved for medical use only for preconfigured specific device combinations. It is not possible
for external manufacturers to integrate their devices into
these networks without the approval and direct cooperation of the iORS provider, and the integration of a foreign
device by the manufacturer is very cost intensive. This is
why a clinic, which has bought an iORS, will be bound to
the manufacturer for a long time and not be free anymore
in its purchase decisions.
In order to overcome all these limitations the idea of
open networked iORS was born, where communication protocols and libraries are open to everyone. In the “SmartOR”
project (2010–2013, 8 partners, a protocol for open network communication in the OR has been
developed, based on a service-oriented architecture SOA
[13]. The technical feasibility of a non-proprietary, modular
integration of medical devices could be shown in general,
but aspects of approval, risk management and legal issues
were not addressed yet. In October 2012 the OR.NET project
continued the work with today more than 90 partners from
all over Germany in order to develop a standardized protocol for device communication, approval strategies and new
methods for risk management of modular OR systems, new
concepts for human-machine-interaction and models for
business partnerships between clinical operators and industrial providers. The protocols are based on the ISO / IEEE
11073 family and are brought into the international standardization process. The project ended in April 2016, but
the non-profit association OR.NET e. V. ( has
been founded in order to further coordinate research and
development activities related to open network communication standards.
3.2 Foot Control and Footswitches
For probably more than 3000 years already humans use
their feet to drive potters-wheels, although the main task
of the feet is to carry the body weight [2]. But if the hands
need to be free for certain tasks, or if a third input is necessary, the feet are a welcome additional input resource
for device handling [12]. For precise manipulation under
limited space in situs handswitches are often suboptimal,
because a release might lead to an unwanted movement of
the instrument which can cause severe damage of tissue.
Footswitches in the medical field are commonly
sold as integral part of a medical device or as additional accessories if they are not mandatory, and only
approved in association with the medical device according to the respective risk class (Medical device directive
93 / 42 / EWG). They have to meet certain technical requirements which arise from the application field OR (e.g electrical safety, protection against liquids).
Footswitches differ in number, kind, color and alignment of input elements. The number of elements ranges
from only one to more than 10 elements. The kind of elements varies according to the needed input format. If a discrete input is needed (on and off), which is the case for many
device functions such as HF coagulation, x-ray imaging or
US cutting, push buttons or digital pedals are used, which
activate a device function as long as they are pushed. For
device functions like milling or drilling, where the speed
control requires a linear input signal to cover a certain range
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
232 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
of a parameter (e. g. position / position or position / speed
proportional), analog foot pedals are used. For device functions which require an increase or decrease of a parameter
value (e. g. zoom or focus) seesaw pedals can be used which
basically consist of two elements and a foot rest as the center
of rotation. For control of functions in 2 degrees of freedom
a joystick or joypad can be used (e. g. activation of the motor
axes of a microscope). The given basic elements of medical
footswitches are shown in Table 2.
The color of footswitch elements is usually not regulated and can be chosen by the device manufacturer.
But for the HF devices it is common sense to produce 2
pedal footswitches with a particular color scheme: yellow
for monopolar functions (left pedal) and blue for bipolar
functions (right pedal). The alignment of elements on a
footswitch is mainly dependent on the number and kind
of elements. Most often footswitch elements are aligned
in line, but for more complex footswitches they can be
aligned in two or more rows, circular or as a combination
of both. Additionally, some footswitches have protective
brackets or covers against unwanted release, and very
often central and side bars are used for a better orientation of the user (Figure 6).
There are only a few footswitches for the control of
several devices. One approach, which basically addresses
the problems related to footswitch shifting and position
changes, is shown in (Figure 7a), where three footswitches
are simply fixed to a solid rack. Another approach is the
angiography footswitch offered by the Siemens Healthcare GmbH (Figure 7b) which can be used to control 8
functions of different devices.
Although devices are interconnected in an IORS
there are no configurable footswitches available for these
systems. The only example of a configurable footswitch is
the iSWITCH offered by Stryker which consist of 2 pedals
for device control of five particular devices and 3 push
buttons to toggle between these devices. This way a lot
more functions can be controlled with only two pedals as
compared to the angiography footswitch presented above
(Figure 7c).
Table 2: Basic elements of medical footswitches (images: Steute Schaltgeräte GmbH).
Push button
Seesaw pedal
Figure 6: Footswitches for medical use (images: Steute Schaltgeräte GmbH, Herga, Bernstein AG, Linemaster Switch Corp., AEI GmbH).
Figure 7: Examples for multi-device-footswitches. a) Solid rack for three different footswitches, seen in the OR1 in the Aqua Clinic Leipzig
(Karl Storz) b) Footswitch for the angiography suite (Siemens) c) iSWITCH configurable footswitch and receiver console (Stryker).
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR The presented approaches have several restrictions.
The solid rack shown in Figure 7a is quite space consuming and heavy, and thus inappropriate for the use under
limited space conditions. For concepts like the angiography footswitch the choice of devices is strongly limited as
well, because the footswitch has to be approved in association with the specific set of devices it shall control.
And the concept is inflexible to function changes, since
a once chosen set of functions cannot be adapted to later
requests. For the iSWITCH concept a visual feedback is
missing to give certainty to the surgeon about the actually chosen function set, and again the choice of devices
is limited to five particular devices only.
4 Requirements Analysis
For the development of a handling concept for a configurable footswitch in the neurosurgical OR a wide range
of requirements has to be considered, which arise from
different aspects. Chapter 4.1 presents workflow observations which have to be done in order to get to know the
roles and working conditions in an OR, and to understand
when and how devices are applied and how the respective
footswitches are used. A sufficient functional set, which fits
the device usage in neurosurgical interventions in general,
and the device usage for the chosen surgical intervention of
a cervical decompression and fusion operation in particular, is defined in chapter 4.2. Further requirements, which
arise from normative guidelines for usability engineering,
are shortly presented in chapter 4.3. A summary of important requirements is given in chapter 4.4.
4.1 Observations and Context of Use
Our field analysis comprises 9 surgical interventions of a
spinal fixation surgery and several other interventions,
such as endoscopic pituitary surgery or open tumor
surgery, and workflows and the device usage have been
documented. Most of the interventions have been carried
out in standing position and took 3 to 4 hours. During the
interventions the room light was dimmed and only the
operational situs was well illuminated. The preparation
of the patient was usually done by the surgical assistant,
and the more complex and demanding tasks were carried
out by a more experienced surgeon. The manipulation
on the patient requires high concentration, and surgeons
react aversely if they are forced to deal with technical or
organizational issues meanwhile.
Most of the surgeons use both feet for footswitch
operation, while some others used their right foot only.
There are footswitches with two or three pedals, however,
only one pedal is used most of the time. E. g. the bipolar
coagulation function has frequently been released by use
of the footswitch, while the monopolar cutting function
was released by some surgeon using the handswitch. The
same is true for the X-ray device, where in the observed
interventions only single shot images were made using
one of the pedals on the respective footswitch. The other
pedal for fluoroscopy control was not used at all. The total
set of control features offered by footswitches is obviously
not needed at all time and pedals are inefficiently used.
Unused pedals are space consuming and one major reason
for the problem of being out of reach to the surgeon.
The device usage in 9 non-navigated spinal fixation
surgeries has been analyzed regarding footswitch usage in
different steps of the workflow. For this purpose a multi
moment analysis has been done, where every 30 seconds
all devices in use have been protocolled and footswitch
handling has been noted. One example is given in Figure 8.
It shows that in the first phase of the intervention, when
the osseous structures of the spine have to be exposed,
monopolar cutting and bipolar coagulation are used in
alternating way. In the phase of screw insertion the C-arm
is mainly used and x-ray images are taken, while in the
phase of laminectomy the milling device is used together
with bipolar coagulation. For the fixation of the vertebrae
again x-ray images are taken.
These operational phases and the described device
functions could be observed in the same or similar way
Figure 8: Sequence of device usage during a spinal fusion operation (multi moment analysis, 30 seconds interval).
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
234 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
in all of the spinal fixation surgeries. For many of the neurosurgical interventions such phases can be identified,
where particular devices are used solely or in combination with other devices. Usually devices for tissue dissection techniques, such as monopolar cutting, milling or US
cutting, are not used in parallel with each other, but in
parallel with device functions such as focus or zoom of the
OR microscope. And for all kinds of interventions the risk
of spontaneous bleeding is always present and bipolar
coagulation therefore is a feature which is used in almost
all phases of a workflow.
Some of the surgeons had to step on a footboard
during the intervention, but the space on a footboard is
limited to 60 cm x 35 cm and there is only little space left
for the positioning of a footswitch. For this reason some
surgeons are used to put two footboards in front of each
other if several footswitches are needed. This way they fall
down from the footboard less frequently, but they still do.
The fact that OR tables do have a bulky column also leads
to spatial conflicts with footswitches, which are even
exacerbated if the C-arm is used.
4.2 Online Survey
A nationwide online survey has been conducted in July
2014 with the aim to get to know habits and problems
related to footswitch handling on the one side, and on the
other side to find out how neurosurgeons would evaluate
the idea of a configurable footswitch and which functions they would expect to control. 351 neurosurgeons
were contacted by email, and 34 neurosurgeons from 13
different university hospitals participated in the survey
(4 female and 30 male subjects; age between 26 and 56
years; 3 clinic directors, 15 consultants, 5 specialists and
11 medical assistants; 1 to 25 years of clinical experience).
The results concerning problems with footswitch handling have already been presented in chapter 2.2. Further
results will be presented now.
The highest number of footswitches in use during
an intervention was given with 5 (n = 2), but most of the
neurosurgeons use three different footswitches in parallel (n = 18). None of the participants had experience with
some kind of configurable footswitch. Most of the surgeons
handle footswitches with both feet in an alternating way
(n = 24), while 9 surgeons only use their right foot and one
only his left foot. In response to the question of the way
the surgeons “find” a footswitch under the OR table and
drapes the haptic approach (palpation) was named most
often (n = 27), followed by the visual approach (look under
the table) with n = 20. The surgeons were asked to make
suggestions for the improvement of footswitch handling,
without being informed about the idea of a configurable
footswitch. They suggested for example “a footswitch
with variable function set according to the actual phase
of the intervention”, “the integration of functions from
the OR microscope”, a “multifunctional panel”, or a
“standard footswitch”. Then the idea of the configurable
footswitch was shortly presented to the participants and
they had to choose functions which they would expect to
find on a configurable footswitch from a given set of 18
functions. The device function “bipolar coagulation” was
named most often, and not only for the general function
set (n = 29), but also to be necessarily available at any
time (n = 27). Further frequently mentioned functions are
milling (n = 24) and x-ray single shot (n = 20). The total list
is presented in Table 3.
4.3 N
ormative Framework for Usability
DIN EN ISO 9241-11:2006-03 describes usability requirements of products. Usability is defined as “the extent to
which a product can be used by specified users to achieve
specified goals with effectiveness, efficiency and satisfaction in a specified context of use”. This means, to which
extent a goal could be reached exactly and completely
(effectiveness), the ratio of effort to degree of goal attainment (efficiency) and how free the user is of impairments
during product usage (satisfaction). DIN EN ISO 60601-1-6
adds the criteria of learnability to the three given criteria
for usability.
DIN EN ISO 9241-210 presents a systematic user
centered approach for the development of interactive
products, based on the needs and requirements of
the users. The process is subdivided into four phases:
comprehension and description of the context of use,
specification of user requirements, creation of design
solutions and testing and evaluation of the design. The
design process allows for a continuous adaption of the
concept in iterative steps: interim results of the diffe­
rent phases are used to adapt the design until it meets
the user requirements.
4.4 Requirements List
According to the presented observations, surveys and
investigations the most important requirements are
listed below. The order does not imply a weighting of
the criteria.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR 235
Table 3: “Which device functions shall be controlled by a configurable footswitch?”.
Table 4: General requirements.
Risk minimization:
Risks due to device operation must be minimized
The configurable footswitch shall be usable in the neurosurgical IOR by the surgeon in an effective, efficient
and satisfying manner
The concept of use shall be easily learned and the system shall be self-descriptive
The system shall support the surgeon in his work, without unnecessary tasks or irrelevant information
Similar information shall be presented in the same manner anywhere in the system
The handling concept shall be as easy as possible
User input must be differentiated reliably and confusion of input elements must be prevented
The concept shall be adaptable to individual needs, e. g. by individual definition of functional pairs
Fault tolerance:
Mistakes in functional choice or handling of the concept shall be corrigible easily and fast
Table 5: Requirements concerning the interaction concept.
System state:
The actual functional set of the configurable footswitch must be visible and clear to the surgeon at any time
Permanent functionality:
Bipolar coagulation must be available at any time and without delay
Footswitch and GUI must give immediate feedback to the user on the effect of an input
Functional set:
The handling concept shall include functions of the HF device, milling device, C-arm, OR microscope,
endoscopic system, navigation system and CUSA. The elements of the footswitch have to provide the
necessary input format (e. g. 0 / 1, analog, two way)
Blind activation:
A reliable operation of the footswitch must be possible without visual control
Functional restriction:
Only functions, which are necessary for the actual intervention, shall be offered during use
Only the surgeon, or somebody under his supervision, must be allowed to change settings of the
footswitch. Automatic changes of the system, e. g. after a certain time period, are inacceptable.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
236 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
Table 6: Requirements concerning hardware design.
Stable standing:
The activation of footswitch elements shall be possible in stable standing
The footswitch shall be operable in standing position, in sitting position and on a footboard
Space-saving design:
The size of the footswitch shall be such that collisions with other equipment are avoided in order to allow
for an optimal body-posture of the surgeon
Position correction:
The position of the footswitch must be adaptable to position changes of the surgeon (without the aid of
Anthropometric design:
The footswitch must be operable by tall men (95 percentile, age 18–69) and small women (5 percentile, age
18–69), wearing OR shoes. The input elements must be designed and aligned such that they can be used
comfortably and safely.
5 Prototype Development
According to the requirement analysis one main function of the configurable footswitch is the activation of
device functions from various devices with different input
format. Additionally it must be displayed to the surgeon
which device functions are actually chosen and which
functions are further available. The navigation through
the available functions in order to change the actual function set must be implemented, with restriction to the abilities of the foot-leg system. The large function pool has
to be structured such that an easy and fast choice and
change is possible. Last but not least, different alignments
and combinations of input elements have to be discussed.
Subchapter 5.1 presents the description of the resulting
solution range.
All these tasks have to be transferred into product
features in order to find adequate realizations, and first
concepts of the system are developed. For comparison
and choice of one concept evaluation criteria are defined
and one concept is chosen for further elaboration, which
is presented in subchapter 5.2. Subchapter 5.3 describes
the first mock-up of the system, which was the base for
a user-centered risk analysis and a wizard-of-Oz experiment. The concept is revised twice and the final prototype
is presented in subchapter 5.4.
5.1 Description of the Solution Range
Figure 9 shows a collection of elementary principles
and characteristics for input and output methods on the
system side and some aspects regarding the foot-leg-system. Using foot movement for system input is generally
possible by force transmission to a mechanical element,
by touch of an input panel or by foot gestures. The input
by touch on a sensor panel or on a touch panel on the
floor is unfavorable because haptic feedback is missing
and blind handling impeded. Foot gestures for digital
input have already been realized in the non-medical
field for device activation [1]. But for the realization of
digital input, which is e. g. necessary for speed control of
a milling device, foot gestures are inappropriate because
precise movements are necessary with missing haptic
feedback. However, the usage of mechanical elements
for foot input is very common. This is because mechanical elements can be located easily due to their shape and
size and haptic feedback provides an immediate system
response. The basic input elements of a footswitch have
already been presented in Table 2.
Specific abilities of the foot-leg-system have to be considered for the design of footswitches, such as the range of
motion and the degrees of freedom of the ankle and knee
joint. Furthermore, stability conditions for upright and
sitting usage are important in order to provide safe handling. In a sitting position the range of motion is bigger
and the stability is higher as compared to the standing
position. This is why footswitch operation shall always be
done with contact to the floor in a standing position, e. g.
through the heel [2].
The structuring of the large function pool could be
done in three different ways. For example, functions can
be structured according to the respective devices, like
seen in the concept of the iSWITCH. This means that the
change of the function set is a change of the device at the
same time. This concept is very close to the present situation with several footswitches in use and would probably
have a high acceptance therefore, but the major problem
is that it is not possible to use two different devices in
parallel, and for devices with only one function the other
pedals are obsolete. Another way of function structuring is according to the usage of a device function in the
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR 237
Figure 9: Solution principles and specifications for the development of a configurable footswitch. Images about foot gestures are taken from
[20] and mechanical elements are taken from [3] and [17].
operational workflow, across devices: pairs of functions
can be defined which are often used in parallel. This
works well if the function pairs fit to the workflow, but
the concept is very inflexible to workflow variations and
the cognitive workload is higher if different function pairs
have to be remembered. The third structuring method is
the presentation of single device functions which can be
chosen from the complete function pool as needed. This
concept would allow for a single-element footswitch,
where the functional assignment is changed for every
device function. Advantages are a high degree of flexibility, easy integration of additional device functions into
the handling concept and the low cognitive workload.
However, a frequent reassignment of the functions might
be necessary depending on the type of intervention.
The combination and alignment of input elements
has a large impact on error rate, simplicity and reliability
of device handling. According to Fitt’s law the accuracy of
element handling is influenced by the ratio between the
distance and the target size [8]. Though, for the design of a
footswitch there is a restriction of the element size due to
the spatial limitations in the OR. Another important aspect
in element handling is the reaction time. According to
Hick’s law a high number of elements leads to an increase
of the reaction time [11]. That is the more input elements
are offered for the selection, the longer the decision which
element to choose. On the other hand, a higher number of
input elements on the footswitch offers the opportunity
of fast device handling in parallel, which in turn is time
saving. The requirement, that bipolar coagulation has
to be available at any time and immediately, which can
only be satisfied by the use of a separate input element
for bipolar coagulation, sets the number of footswitch
elements to a minimum of two. Additional elements will
be necessary for the release of further functions and for
the change of the function set. The analysis of the input
format for the device functions from Table 3 shows that
three types of elements are appropriate: a pedal for discrete input (all device functions which are switched on
and off), a pedal for analog input (speed control of the
milling device) and a seesaw pedal for two way change of
a parameter (e. g. zoom and focus of the OR microscope).
Since analog input can be converted into a discrete signal
through the definition of a threshold the analog pedal can
be used for both input formats. The change of the function
set can be done either by one or two elements, which are
exclusively used for toggling between functions / function
pairs and an element for the confirmation of the change, or
by using only one additional element which enables some
kind of “selection mode” where the elements for function
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
238 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
release are used to toggle. In the latter concept function
release and function change is done by use of the same
elements which might cause a confusion of the user, but
on the other hand fewer elements are necessary, which is
space saving and enhances the handling accuracy.
5.2 D
evelopment of First Concepts and
Concept Selection
Different solutions for the basic functions “device release”,
“change of mode” and “toggling” have been worked out
and combined in four concepts for the design of the configurable footswitch (Table 7). The pedal for bipolar coagulation is colored in blue and always on the right side, as
it is on all footswitches of the HF devices. Red color is used
to indicate elements which are involved in a change of the
function set. In the first two concepts the pedals are only
used for device release and function change is done by use
of additional elements, while concepts 3 and 4 share these
elements. Toggling is done by use of a joystick (concept 1),
by touch of two sensor-bars (concept 2), and through activation of pedals (concepts 3 and 4). In concept 1 and 2
only pairwise function change is possible, while concepts
3 and 4 offer single and pairwise change.
These four concepts have been compared with each
other, based on the following criteria: Time for function
change, individualization, simplicity, stable standing,
need of space, physical load, self-descriptiveness and
the risk of unwanted modus changes. The concept with
the maximum score is concept 4. It gets the highest score
for the criteria “individualization”, “simplicity”, “need of
space” and “time for function change” as compared to the
other concepts. Only few ratings are given for the criteria
“stable standing” and “risk of unwanted modus change”,
because the user has to lift one foot completely in order to
kick against the upper elements (risk of unstable stand)
and the upper elements might be activated by accident
during device release if the foot is moved too far forward
(risk of unwanted modus change). These limitations are
considered for further development.
5.3 Prototype Development
The prototype consists of two sub elements: the footswitch
as input interface and the GUI as visual output on the
central working station.
The footswitch has been built in the first step in form
of a non-electric mock-up (Figure 10 left), which could be
used for some pre-tests regarding size and operability. The
GUI has been implemented with Blend for Visual Studio,
as a mock-up of a user interface with no underlying function, but offering parts of the operating logic (Figure 10
right). This GUI mock-up could then be analyzed regarding
risks in human machine interaction with a model-based
Table 7: Concepts for the design of a configurable footswitch.
Concept 1
Concept 2
Concept 3
Concept 4
Figure 10: Preliminary design studies of the configurable footswitch and its GUI. Left: Wood and paper based footswitch mock-up, middle:
paper draft of the GUI, right: GUI mock-up, built with Blend for Visual Studio.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR human risk analysis method [14], and the design and the
operating logic were revised according to the results from
the analysis.
Then both dummies were used within a Wizard-of-Oz
experiment [6], where four neurosurgeons had to perform
certain tasks with the system and evaluate their user
experience in a questionnaire. The results of this experiment were then considered for the production of the final
5.4 Final Prototype
The final prototype of the footswitch consists of 3 elements for device release and 2 elements only for function
change. A blue pedal on the right is used solely for bipolar
coagulation. In the middle there is a seesaw pedal with
a central footrest for two-way change of function parameters. The left pedal is colored in yellow and allows for
both, analog and discrete input. Two black push button
elements are positioned on the upper left and right side,
and two lateral stops serve a better orientation of the foot
(Figure 11). All footswitch elements are approved. The
technical connection to the OR network is done wirelessly
through a receiver which is connected to a “Raspberry Pi”
single board minicomputer.
The GUI prototype is integrated into the central
working station. It consists of several subsections: A
model of the footswitch with the actual function set in
the center of the screen, a selection of single device functions for the yellow pedal and the grey seesaw pedal on
the lower edge, a selection of function pairs (‘presets’)
on the right side, and a small model of the two configurable pedals with the actual function set on the lower left
edge (Figure 12). The black push button elements of the
footswitch are indicated by two rectangle fields above the
footswitch elements in the center of the screen, filled with
the labels “Single Selection” and “Presets”.
Figure 12 shows the so called “working mode”, where
the actual function set (yellow: cutting with US bone
Figure 11: Final prototype of the configurable footswitch. Left: CAD model; middle: motion direction for activation; right: physical prototype.
Figure 12: Functional lab type of the GUI of the configurable footswitch.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
240 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
knife, grey: microscope zoom) is ready for device release.
Changing of the function set can be done either separately
for each configurable pedal (“Single Selection”) or for
both pedals at once (“Presets”), through activation of a
black push button element on the footswitch. E. g. if the
left black push button is activated the system changes its
state from “Working Mode” to “Selection Mode – Single”,
and the label of the upper left rectangle changes from
“Single Selection” to “OK” (Figure 13 left). To change the
function of the yellow pedal it has to be activated once,
and a fly out field appears with all available device functions (Figure 13 right). Toggling through the functions is
done by repeated activation of the yellow pedal and works
from the left to the right side, and then again from the left
side. The confirmation of the function selection is done by
activation of the left black push button element again, as
indicated by the label “ok”.
The selection of function pairs (“presets”) is done likewise by activation of the upper right push button element.
The mode changes into “Selection Mode – Presets”, and
toggling through the available presets is done bidirectional using both parts of the seesaw pedal. Activation of
the upper right push button ends the selection mode and
confirms the function change.
The blue pedal for bipolar coagulation is inactive
if the system is in any kind of selection mode. This way
an unwanted release of the function bipolar coagulation
during the selection process is inhibited.
6 Usability Evaluation
The usability evaluation of the handling concept has been
done in a user-centered experiment, where certain tasks
had to be performed by surgeons in an experimental OR
environment. The tasks were designed on the basis of
the device usage of a navigated posterior cervical spine
operation (Figure 16) and had to be performed in two
settings: the configurable footswitch versus the conventional setting with 4 different footswitches. The subjects
got a standard introduction to the usage of the configurable footswitch by the investigator and were free to try
it out until they felt safe with the handling. They were
asked about their emotional condition before and after
every experimental cycle, and the answers were classified
according to the shackle scale (0 = totally relaxed until
20 = unbearably stressed). During the experimental cycles
Figure 13: Single selection of a device function. Left: State after activation of the upper left element. Yellow and grey pedal can be selected
now by single activation. Right: Fly out field with device functions for the yellow pedal.
Figure 14: Left: Fly out field with device functions for the grey seesaw pedal. Right: Selection mode for function “presets”.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR the investigator issued instructions and made announcements about spontaneously occurring and invisible bleeding at different numbers. An assistant undertook the task
of the sterile nurse to take and pass instruments from and
to the surgeon.
The experiments were filmed from two positions:
one camera giving a good overview over the total setup
and a second camera with focus on footswitch handling
(Figure 15). Additionally a screen recording was used to
save the interaction with the GUI of the configurable
footswitch. After the last experimental cycle the subjects
had to fill out a questionnaire, containing 53 statements,
which had to be rated by the subjects from “I fully agree”
in 4 steps to “I fully disagree” and space for open statements, and they had to do a NASA-TLX test with the conventional setting as reference [10]. Some surgeons also
started a discussion about advantages and disadvantages
of the new system, which could be transliterated later on.
Figure 15: Experimental setup within the integrated demonstrator OR. The conventional setting with 4 different footswitches is shown in
the upper right picture. Red circles: Two HD cameras used for documentation of the footswitch handling; Yellow circle: Screen of the central
working station with the GUI of the configurable footswitch.
Figure 16: Experimental setups: A) monopolar cutting of a shamrock and bipolar coagulation at numbers announced by the investigator.
B) navigated introduction of a pointer (simulation of drilling) into canals for pedicle screws, documentation of the pointer position,
simulated x-ray image acquisition, bipolar coagulation at announced numbers. C) Cutting of a line (A–B and C–D) using the US bone knife,
while bleeding is announced several times at different numbers in the background (black dots under the arrows, only readable when
focused), focusing with the microscope between foreground and background, taking snapshots of the microscope field of view.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
242 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
9 surgeons (5 neurosurgeons and 4 orthopedic surgeons) participated in the experiment, and the overall
duration was about one hour (including 30 minutes for
the experimental cycles). All surgeons were familiar with
posterior cervical decompression and fusion operations.
The chronological sequence of the user tests is shown in
Table 8: A double cross over design was used (4 experimental cycles) in order to handle the small number of subjects and to evaluate potential learning effects later on.
The assignment of a subject to group 1 or 2 has been done
in an alternating way.
After the end of the user tests both camera perspectives and the screen loggings have been synchronized and
fused into one video for each subject. These videos served
as the base for the subsequent data acquisition, which is
done with respect to the usability criteria effectiveness,
efficiency, learnability and satisfaction.
The evaluation regarding effectivity is based on
the documentation in number and kind of problems in
footswitch handling and always referring to the actual
number of activations, since there are variations in activation numbers between the conventional and the new
setting due to technical problems and due to mistakes in
task fulfillment.
The evaluation of the efficiency is basically done
through comparison of durations of the different experiments. Unfortunately it is not possible to compare the
total duration of each experiment, because too many
disturbing factors lead to a contamination of the data.
For this reason only the interval between an announcement by the investigator of a spontaneous bleeding and
the beginning of the coagulation by the surgeon, and the
interval between the announcements by the investigator
that bleeding is stopped and the continuing cutting by the
surgeon is measured. This way interruptions and variations in time between announcements by the investigator
can be filtered for experiments A and C. Experiment B
cannot be considered for efficiency evaluations, because
the fictive drilling process lead to high variations in time,
which impedes a comparison between both setups.
For effectivity and efficiency evaluations only the
third and fourth experimental cycle are considered,
because strong learning and carry-over effects between
the first two cycles have to be expected. These are now
regarded as a phase of familiarization with the setup and
the experimental tasks, and only cycles three and four are
evaluated using statistical methods for cross over design
studies [19].
Learnability is investigated through comparison of
handling errors and handling durations across the experimental cycles. The user satisfaction is analyzed based on
their ratings and statements in the questionnaires and in
7 Results
In this chapter the results of the usability evaluation are
presented. According to the definition of usability the
evaluation is subdivided into three parts addressing effectiveness, efficiency and learnability.
7.1 Effectiveness
For cycle three and four there could be observed 9 handling errors (total number of activations is 272) by the use
of the conventional setting, while 28 handling errors (total
number 359) occurred using the configurable footswitch
(CFS), which means that the usage of the CFS lead to 5 %
more handling errors. The statistical evaluation for a cross
Table 8: Chronological sequence of the user experiments.
Group 
5 subjects
Group 1
 subjects
Start Introduction,
Explanation of
tasks A, B and C
Presentation &
Explanation of
explanation of
tasks A, B and C the CFS
Presentation &
explanation of
the CFS
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR over design showed that a carry-over effect between both
cycles is probable (Tw = 1,911 > t(7;0.95) = 1,895), and this is
why only results from cycle three may be considered and
evaluated as a parallel group study. Here the test statistic
is lower than the t value (T = 0.95 < t(7;0.95) = 1,895) and an
influence of CFS usage to handling errors is not significant
to the two-tailed 10 % level [9].
All handling errors are classified according to several
error types and causes. The error types for the CFS and of
the conventional setting differ to a certain extent due to
the varying operating options of both systems. The breakdown of errors according to error types is given in Table 9.
It shows that the most frequent error in CFS handling was
the accidental activation of a second pedal (n = 14), followed by futile device activation (n = 6) due to the system
state “selection mode” (where device release is inhibited),
which was activated through accidental kicking against
the black elements. For the conventional setting the most
frequent error was the activation of a wrong pedal on the
right footswitch (n = 5), followed by the activation of a
pedal on the wrong footswitch (n = 3).
In cycles three and four 79 changes of the functional
set of the CFS were done all in all, and only 4 mistakes
could be observed.
7.2 Efficiency
Regarding differences in the duration of single operations
in tasks A and C the test statistic for cross-over studies
does not identify a carry-over effect between cycles three
and four (Tw_A = 1.10 and Tw_C = 1.40 < t(7; 0.95) = 1.895).
Both cycles are considered. The results for task A and C
differ: While in task A the differences are not significant
(t = 1,741 < t(7; 0.95) = 1.895), in task C the duration for
single operations was significantly shorter when the CFS
was used (t = 3.056 > t(7; 0.975) = 2,365). Furthermore, the
number of palpation of the footswitch and of looking at
the footswitch was counted, and it showed that for the
conventional setting it was twice as high for both.
7.3 Learnability
Learning effects are evaluated regarding effectiveness in
footswitch handling, efficiency for single handling operations, and effectiveness in changing of the functional
set of the CFS. All evaluations are based on a comparison
between cycles one or two with three or four.
Regarding the effectiveness in footswitch handling
no learning effect can be seen, but even an increase
of handling errors for both systems in the respective
second cycle.
Regarding efficiency a significant learning effect can
be observed with the usage of the CFS for task A (t = 2.05 >
t(0.95, 16) = 1,746), where the mean duration of single handling operations was reduced by 24 % in the second cycle.
For task C no significance could be shown (t 0.95), although
8 of 9 subjects were faster or equally fast in the second
cycle. For the conventional setting no significant gain in
handling velocity can be observed for task A, while in task
C handling was about 17 % faster in the second cycle.
Table 9: Operating errors observed during usage of the configurable footswitch (CFS) and of the conventional setting. (green color marks
errors resulting from hardware design).
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
244 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
For the change of the function set of the CFS a clear
learning effect can be observed (t = 2.47 > t(0.975 = 2.228):
while in the first cycle 22 % of faulty operations occurred,
the number reduced to only 8 % in the second cycle, where
4 subjects even made no mistakes at all (Table 11).
The temporal occurrence of errors also indicates a learning effect, because 66 % of all errors occur in the first third of
a cycle, while no errors at all are observed in the last third.
7.4 User Satisfaction
The evaluation of the emotional status (taken before and
after each cycle) according to the Shackle scale shows
that all subjects were somewhat to very relaxed during the
whole experiment. The highest values (6 = still acceptable)
were taken before the experimental cycles started and
reduced after the first cycle for most subjects. After experimental cycles with the CFS the values reduced on average
and increased again when the conventional setting was
used. According to the NASA-TLX test the mental demand
for the handling of the CFS is higher than with the conventional setting, but for physical demand and frustration the
CFS got better results than the conv. setting (Table 12). For
the other criteria the results were more or less balanced. In
the questionnaire 5 of 9 subjects stated that they generally
like the user interaction concept of the CFS and 8 subjects
think that it offers all necessary functions. 7 subjects think
that the number of input elements is sufficient and all subjects stated that the GUI gives an appropriate overview of
available functionalities. For 5 subjects it was problematic
to find the input elements without visual contact, especially for the black button elements. But problems with
instability during one-leg-stand could not be observed.
Table 10: Comparison of the mean duration for single handling operations in task A and C.
Table 11: Occurrence of errors in function change of the CFS. Left: comparison between first and second cycle. Right: combined presentation
of the temporal occurrence of errors for both cycles.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR 245
Table 12: Selected results from the NASA-TLX test. The workload for CFS usage has been calculated through multiplication of scale and
weighting factors, and the difference to the workload with the conventional setting is shown.
8 Discussion and Outlook
Although from a statistical point of view no significant
increase of handling errors could be shown it still must
be supposed that there is an effect, with regard to the
higher percentage of handling errors during CFS usage.
However, a closer look at the types of errors in CFS usage
shows that 80 % have their cause in a suboptimal design
of the footswitch hardware (e. g. the accidental activation
of a second pedal Table 9), and only 6 errors occurred due
to general problems with the handling concept, which is
even less as compared to the conventional setting where
8 errors related to problems with the handling concept
occurred. This is a promising result for the general usability
of the handling concept of the configurable footswitch and
implies that enhancements of the hardware design have the
potential to reduce the number of handling errors. The most
frequent problem, that two pedals are hit at once, could
be solved by central bars between the pedal as proposed
by several subjects after the experiments. The distance
between the black push button elements and the pedals
has to be enlarged in order to prevent unwanted activation
of the black push button elements, which enables the selection mode. Furthermore, it is recommended to either place
the black push button elements above the pedals or to the
sides in order to avoid unwanted release during pedal activation. Regarding the increase of the efficiency for task C
it can be assumed that users benefit most from the usage
of a CFS if several devices are used in parallel, and especially if the hands are used in the conventional setting
for device handling (e. g. for the microscope). The results
for handling errors and learnability during the process
of function change show good prospects and imply that
self-descriptiveness and operating logic of the system are
supportive to the user. Regarding the mental demand the
subjects have contrary opinions: 4 think the workload due
to the mental demand is lower with the CFS, and 4 think
it is much higher. This result could be expected, since the
surgeons were confronted with an entire new handling
concept and had only little time for familiarization. To sum
up, the handling concept was rated very positively by the
surgeons and most of them could imagine that, after some
revisions, such a system will establish in the future. The still
existing technical problems with the first prototype have to
be solved and the concept shall be more flexible regarding
the assignment of functions to pedals.
There are some limitations of this study. The number
of subjects is rather low, and for a statistical evaluation
with methods for a cross-over-design some requirements,
such as wash-out-periods between the cycles and a randomized assignment of subjects to both groups, could
not be fulfilled. Furthermore, interruptions due to system
crashes and problems with the black push button elements for mode selection (elements did not react or activated double) were disturbing. And though the number
and kind of devices used in the experiments was representative, the high frequency of device and instrument
changes was unrealistic.
The new system of a configurable footswitch competed
against the conventional device specific footswitches all
surgeons are familiar with from daily clinical routine. But
despite these testing conditions and despite all given limitations it can be resumed that the CFS fulfilled most of
the defined requirements and was rated very positively
by the representative user group. Further revisions of the
concept are in progress, which include the transfer of the
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
246 J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR
concept to other surgical disciplines such as orthopedics
and ear-nose-throat surgery.
Some general aspects of the presented work might
be interesting for further safety critical applications in
medicine or other fields, where several footswitches are
used during complex processes. The assignment of safety
critical functions and of less critical functions to different
footswitch elements can be reasonable for any system,
where a classification of functions into different risk levels
is possible. In the whole surgical field a reasonable classification of device functions might consider, whether a
device function causes tissue damage or not, while classification criteria in other safety critical fields might differ
considerably dependent on the application. Additionally
the fact that there can be functions which have to be available at any time, such as bipolar coagulation in our application, can also hold true for other systems and might
even constitute a major requirement for the safety of such
systems. Furthermore it has been seen that users, who are
strongly focused on their task, often don’t pay much attention to their working conditions, although there might be
a high potential for improvement of HMI. A view from the
outside might be necessary in order to overcome operational-blindness and enhance workflow and safety.
Acknowledgements: This work was funded by the German
Federal Ministry of Education and Research (BMBF) in the
context of the OR.NET project (16KT1203). The author thanks
Olivia Thoma for her support, who did her master thesis in
this topic, and to all surgeons from the Neurosurgical and
Orthopedic Clinic, who took their time to participate in the
user test. Furthermore, the close and valuable cooperation
with the Department of Neurosurgery at the RWTH Aachen
University Hospital over the last years shall be mentioned.
And we are thankful to the Steute Schaltgeräte GmbH, who
provided hardware components for the footswitch prototype.
[1] Alexander, J., Han, T., Judd, W., Irani, P., & Subramanian, S.
(2012). Putting your best foot forward: investigating real-world
mappings for foot-based gestures. In Proceedings of the
SIGCHI Conference on Human Factors in Computing Systems
(pp. 1229–1238).
[2] Bandera, J. E., Muntzinger, W., & Solf, J. J. (1989). Auswahl
und Gestaltung von ergonomisch richtigen Fußstellteilen.
Schriftenreihe der Bundesanstalt für Arbeitsschutz: 590, 1.
Bremerhaven: Wirtschaftsverlag.
[3] Baumann, K., & Lanz, H. (1998). Mensch-MaschineSchnittstellen elektrischer Geräte (1st ed.). Berlin,
Heidelberg: Springer Verlag.
[4] Blaar, M. (2013). Analyse nutzungsorientierter Risiken im
Rahmen einer Workflowoptimierung im neurochirurgischen
Operationssaal (Dissertation). Rheinisch-Westfälische
Technische Hochschule Aachen, Aachen.
[5] Börm, W., Meyer, F. (2009). Spinale Neurochirurgie –
Operatives Management von Wirbelsäulenerkrankungen.
Stuttgart, Germany: Schattauer Verlag.
[6] Dahlbaeck, N., Jönsson, A., Ahrenberg, l. (1992). Wizard of Oz
studies – why and how. In: Proceedings of the 1st international
conference on Intelligent user interfaces (IUI ’93), ACM,
New York, NY, USA.
[7] Dell’Anna, J., Janß, A., Blaar, M., Höllig, A., Clusmann, H.,
Radermacher, K. (2014) Analysis of User-Induced Risks in
the Neurosurgical Operation Room. In Proceedings of the
5th International Conference on applied Human Factors and
Ergonomics 2014, pp. 352–358.
[8] Fitts, P. M. (1954). The information capacity of the human motor
system in controlling the amplitude of movement. Journal
of Experimental Psychology, 47(6), 381–391. doi:10.1037/
[9] Gaus, W., Muche, R. (2014). Medizinische Statistik –
Angewandte Biometrie für Ärzte und Gesundheitsberufe.
Stuttgart, Germany: Schattauer Verlag.
[10] Hart, S. G., Stavenland, L. E. (1988). Development of NASA-TLX
(Task Load Index): Results of empirical and theoretical
research In Human Mental Workload, In: Peter A. Hancock and
Najmedin Meshkati, Editor(s), Advances in Psychology, Vol. 52,
pp. 139–183, North-Holland, Elsevier.
[11] Hick, W. E. (1952). On the rate of gain of information.
Quarterly Journal of Experimental Psychology, 4(1), 11–26.
[12] Hölscher, U., Laurig, W., & Müller-Arnecke, H. W. (2008).
Prinziplösungen zur ergonomischen Gestaltung von
Medizingeräten (2. Aufl.). Dortmund: BAuA.
[13] Ibach, B. (2011): Konzeption und Entwicklung einer
serviceorientierten Integrationsarchitektur für die Vernetzung
von Medizinprodukten im Operationssaal. In: S. Leonhardt,
K. Radermacher & T. Schmitz-Rode (ed.): Aachener Beiträge zur
Medizintechnik (ISBN 978-3-8440-0298-0), Shaker, 2011, pp. 166.
[14] Janß, A., and Radermacher, K. (2014). Usability First. Modellbasierter
Ansatz zur nutzungsorientierten Risikoanalyse von Medizinprodukten
[Usability first. Model-based approach for the use-oriented risk
analysis of medical devices]. Bundesgesundheitsblatt,
Gesundheitsforschung, Gesundheitsschutz, 57(12), 1384–1392.
[15] Liehn, M., Steinmüller, L., & Döhler, J. (2011). OP-Handbuch:
Grundlagen Instrumentarium OP-Ablauf (5th ed.).
Berlin Heidelberg, Germany: Springer Verlag.
[16] Matern, U., Koneczny, S., Scherrer, M., & Gerlings, T. (2006,
November 24). Arbeitsbedingungen und Sicherheit am
Arbeitsplatz OP. Deutsches Ärzteblatt, pp. 3187–3192.
[17] Neudörfer, A. (2011). Gestaltung ergonomiegerechter
Maschinen. In A. Neudörfer (Ed.), Konstruieren
sicherheitsgerechter Produkte (pp. 487–528).
Berlin, Heidelberg. Germany: Springer.
[18] Nowatschin, S. (2009). CIO – Computer Integrated Operating
Room – Neue Konzepte und Systeme für einen ComputerIntegrierten Operationssaal. Dissertation, TU-München, 2009.
[19] Schulgen, G., Schumacher, M. (2008). Methodik klinischer
Studien: Methodische Grundlagen der Planung, Durchführung
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
J. Dell’Anna et al., A Configurable Footswitch Unit for the Open Networked OR und Auswertung (Statistik und ihre Anwendungen).
Springer-Verlag, Berlin-Heidelberg-New York 2008, 3.
Auflage, S. 305–308 ISBN 978-3-5408-5135-6.
[20] Scott, J., Dearman, D., Yatani, K., & Truong, K. N. (2010).
Sensing foot gestures from the pocket. In Proceedings of the
23nd annual ACM symposium on User interface software and
technology (pp. 199–208).
[21] Van Veelen, M. A., Snijders, C. J., Van Leeuwen, E., Goossens,
R. H. M., & Kazemier, G. (2003). Improvement of foot pedals
used during surgery based on new ergonomic guidelines.
Surgical endoscopy, 17(7), 1086–1091. doi:10.1007/s00464002-9185-z.
Jasmin Dell’Anna
RWTH Aachen University,
Chair of Medical Engineering,
Aachen, Germany
Jasmin Dell’Anna was born in Desio, Italy, in 1982. She received
the Dipl.-Ing. degree in mechanical engineering from the Cologne
University of Applied Sciences, Cologne, Germany, in 2006 and a
M.Sc. degree in Biomedical Engineering from the RWTH Aachen
University, Aachen, Germany, in 2009. In 2010, she joined the
Chair of Medical Engineering at RWTH Aachen University, Aachen,
Germany, as a member of the scientific staff. She was working on
the development of new handling concepts for the open integrated
OR within the OR.NET project, and on new methods for risk
management and approval for such modular medical systems.
Armin Janß
RWTH Aachen University,
Chair of Medical Engineering,
Aachen, Germany
Armin Janß was born in Cologne, Germany, in 1974. He received the
Dipl.-Ing. degree in electrical engineering from the RWTH Aachen
University, Aachen, Germany, in 2005 and the Dr.-Ing. degree
from the RWTH Aachen University, Aachen, Germany, in 2016.
He has been a member of the scientific staff at the Chair of Medical
Engineering at RWTH Aachen University, Aachen, Germany since
2006 and has been the leader of the group for “Risk Management,
Ergonomics and Usability” since 5 years. In the OR.NET project he
was the head of the working groups “Human Machine Interaction”
and “Risk Management and Approval”.
Hans Clusmann
RWTH Aachen University,
Department of Neurosurgery,
Aachen, Germany
Hans Clusmann was born in 1965, graduated from medical school at
the University of Cologne in 1993. He obtained the doctoral degree
in 1996, the second thesis (Habilitation) in 2004. From 1994,
he underwent training in neurosurgery at Bonn University. He
passed the neurosurgical board exam in 2001, became staff member
in Bonn until 2010, when he was elected professor and chairman
of the Department of Neurosurgery at RWTH Aachen University.
HC covers a broad clinical spectrum in cranial, skull base, pediatric,
and complex spinal surgery. However, his scientific work has been
primarily associated with modern neurosurgical and imaging
techniques as well as basic mechanisms and outcome analyses
in epilepsies. In the OR.NET project he was a clinical partner and
worked on the development of new handling concepts for the open
integrated neurosurgical OR.
Klaus Radermacher
RWTH Aachen University,
Chair of Medical Engineering,
Aachen, Germany
Klaus Radermacher was born in 1964. He received the Dipl.-Ing.
degree in Mechanical Engineering from the Technische Universitaet
Darmstadt, Germany, in 1989, the Physikum in Human Medicine
at the Mainz University, Germany, in 1990 and a doctoral degree
(Dr.-Ing.) from the Faculty of Mechanical Engineering, RWTH Aachen
University, Germany in 1999. From 1988 to 1990 he was engineering
associate at the Institute for Human Factors at Darmstadt University
and research associate in biomedical engineering of the Research
Association for Biomedical Engineering (FGBMT e. V.) in Aachen from
1990 to 2001. From 2001 to 2005 he was cofounder and CEO of the
SurgiTAIX AG, Herzogenrath, Germany and Senior Researcher at the
Chair of Applied Medical Engineering, Medical Faculty, RWTH Aachen
University. Since 2005 he is Full Professor and Lecturer in Medical
Engineering and head of the Chair of Medical Engineering of the
Faculty of Mechanical Engineering (
at the Helmholtz-Institute for Biomedical Engineering, RWTH Aachen
University, and Director of the Institute. In the OR.NET project he was
the project coordinator.
Brought to you by | University of Florida
Download Date | 10/26/17 7:00 AM
Без категории
Размер файла
47 964 Кб
2016, 0031, icoml
Пожаловаться на содержимое документа