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.
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 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].
Use of different conventional device specific footswitches in the neurosurgical OR.
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 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.
Standard equipment of a neurosurgical OR [15].
| Medical device | Function | Input device |
|---|---|---|
| Endoscopic system | Minimal invasive diagnosis and treatment | handswitch |
| 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 | footswitch |
| 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 |
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 intervention.
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.
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.
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.
2.2 Footswitches 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.
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.
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).
Footswitch situation at present: need for improvement.
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 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, www.smartor.de) 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. (www.ornet.org) 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 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.
Basic elements of medical footswitches (images: Steute Schaltgeräte GmbH).
 |
 |
 |
 |
 |
 |
| Push button | Pedal | Plate | Seesaw pedal | Joypad | Joystick |
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).
Footswitches for medical use (images: Steute Schaltgeräte GmbH, Herga, Bernstein AG, Linemaster Switch Corp., AEI GmbH).
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.
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).
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).
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.
Sequence of device usage during a spinal fusion operation (multi moment analysis, 30 seconds interval).
These operational phases and the described device functions could be observed in the same or similar way 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.
“Which device functions shall be controlled by a configurable footswitch?”.
 |
General requirements.
| Risk minimization: | Risks due to device operation must be minimized |
| Usability: | The configurable footswitch shall be usable in the neurosurgical IOR by the surgeon in an effective, efficient and satisfying manner |
| Learnability: | The concept of use shall be easily learned and the system shall be self-descriptive |
| Adequacy: | The system shall support the surgeon in his work, without unnecessary tasks or irrelevant information |
| Consistency: | Similar information shall be presented in the same manner anywhere in the system |
| Simplicity: | The handling concept shall be as easy as possible |
| Accuracy: | User input must be differentiated reliably and confusion of input elements must be prevented |
| Customizability: | 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 |
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 |
| Feedback: | 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 |
| Controllability: | 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. |
Requirements concerning hardware design.
| Stable standing: | The activation of footswitch elements shall be possible in stable standing |
| Flexibility: | 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 hands) |
| 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. |
4.3 Normative Framework for Usability Engineering
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 different 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.
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.
![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].](/document/doi/10.1515/icom-2016-0031/asset/graphic/j_icom-2016-0031_fig_059.jpg)
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 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 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 Development 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.
Concepts for the design of a configurable footswitch.
| Concept 1 | Concept 2 | Concept 3 | Concept 4 |
|---|---|---|---|
 |
 |
 |
 |
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 human risk analysis method [14], and the design and the operating logic were revised according to the results from the analysis.
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.
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 prototype.
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.
Final prototype of the configurable footswitch. Left: CAD model; middle: motion direction for activation; right: physical prototype.
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”.
Functional lab type of the GUI of the configurable footswitch.
Figure 12 shows the so called “working mode”, where the actual function set (yellow: cutting with US bone 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”.
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.
Left: Fly out field with device functions for the grey seesaw pedal. Right: Selection mode for function “presets”.
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 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.
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.
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.
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.
Chronological sequence of the user experiments.
Start  End |
|||||||
|---|---|---|---|---|---|---|---|
| Group 1 4 subjects |
Introduction, Explanation of tasks A, B and C |
 |
Presentation & explanation of the CFS |
 |
 |
 |
Questionnaire and NASA-TLX |
| Group 2 5 subjects |
Introduction, Explanation of tasks A, B and C | Presentation & explanation of the CFS |
 |
 |
 |
 |
Questionnaire and NASA-TLX |
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 the NASA-TLX.
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 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).
Operating errors observed during usage of the configurable footswitch (CFS) and of the conventional setting. (green color marks errors resulting from hardware design).
 |
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.
Comparison of the mean duration for single handling operations in task A and C.
 |
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).
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.
 |
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.
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 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.
About the authors
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 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 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 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 (www.meditec.rwth-aachen.de) 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.
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.
References
[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).10.1145/2207676.2208575Search in Google Scholar
[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.Search in Google Scholar
[3] Baumann, K., & Lanz, H. (1998). Mensch-Maschine-Schnittstellen elektrischer Geräte (1st ed.). Berlin, Heidelberg: Springer Verlag.10.1007/978-3-642-58802-0_1Search in Google Scholar
[4] Blaar, M. (2013). Analyse nutzungsorientierter Risiken im Rahmen einer Workflowoptimierung im neurochirurgischen Operationssaal (Dissertation). Rheinisch-Westfälische Technische Hochschule Aachen, Aachen.Search in Google Scholar
[5] Börm, W., Meyer, F. (2009). Spinale Neurochirurgie – Operatives Management von Wirbelsäulenerkrankungen. Stuttgart, Germany: Schattauer Verlag.Search in Google Scholar
[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.10.1145/169891.169968Search in Google Scholar
[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.Search in Google Scholar
[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/h0055392.Search in Google Scholar
[9] Gaus, W., Muche, R. (2014). Medizinische Statistik – Angewandte Biometrie für Ärzte und Gesundheitsberufe. Stuttgart, Germany: Schattauer Verlag.Search in Google Scholar
[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.10.1016/S0166-4115(08)62386-9Search in Google Scholar
[11] Hick, W. E. (1952). On the rate of gain of information. Quarterly Journal of Experimental Psychology, 4(1), 11–26. doi:10.1080/17470215208416600.Search in Google Scholar
[12] Hölscher, U., Laurig, W., & Müller-Arnecke, H. W. (2008). Prinziplösungen zur ergonomischen Gestaltung von Medizingeräten (2. Aufl.). Dortmund: BAuA.Search in Google Scholar
[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.Search in Google Scholar
[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. doi:10.1007/s00103-014-2063-8.Search in Google Scholar PubMed
[15] Liehn, M., Steinmüller, L., & Döhler, J. (2011). OP-Handbuch: Grundlagen Instrumentarium OP-Ablauf (5th ed.). Berlin Heidelberg, Germany: Springer Verlag.10.1007/978-3-642-16845-1Search in Google Scholar
[16] Matern, U., Koneczny, S., Scherrer, M., & Gerlings, T. (2006, November 24). Arbeitsbedingungen und Sicherheit am Arbeitsplatz OP. Deutsches Ärzteblatt, pp. 3187–3192.Search in Google Scholar
[17] Neudörfer, A. (2011). Gestaltung ergonomiegerechter Maschinen. In A. Neudörfer (Ed.), Konstruieren sicherheitsgerechter Produkte (pp. 487–528). Berlin, Heidelberg. Germany: Springer.10.1007/978-3-642-19189-3_7Search in Google Scholar
[18] Nowatschin, S. (2009). CIO – Computer Integrated Operating Room – Neue Konzepte und Systeme für einen Computer-Integrierten Operationssaal. Dissertation, TU-München, 2009.Search in Google Scholar
[19] Schulgen, G., Schumacher, M. (2008). Methodik klinischer Studien: Methodische Grundlagen der Planung, Durchführung und Auswertung (Statistik und ihre Anwendungen). Springer-Verlag, Berlin-Heidelberg-New York 2008, 3. Auflage, S. 305–308 ISBN 978-3-5408-5135-6.Search in Google Scholar
[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).10.1145/1866029.1866063Search in Google Scholar
[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/s00464-002-9185-z.Search in Google Scholar PubMed
© 2016 Walter de Gruyter GmbH, Berlin/Boston
