Projekte

Gewinnen Sie einen schnellen Überblick unserer laufenden und abgeschlossenen Forschungs- und Entwicklungsprojekte. Die Projekte sind nach Forschungsgruppen sortiert. Es ist zu beachten, dass verschiedene Projekte gemeinsam von mehreren Forschungsgruppen am HuCE bearbeitet werden.

WiseSkin for tactile prosthetics
Intuitive and reliable representation of lung ventilation by electrical impedance tomography
VoiSee - A Portable Electronic Vision Aid for Larger Pictures and Better Contrast, in Particular for Patients with Macular Degeneration
Cardiac Simulator
Development of intra-vaginal sensors to measure pelvic floor muscle activity
Determination of the Transit Time through the Intestine of Cows with Cecum Dilatation-Dislocation
Fast Data Processing for Multi-Electrode-Arrays
A physiologic robot reactor system to simulate in vivo conditions

WiseSkin for tactile prosthetics

Volker M. Koch, Jörn Justiz, Ozan Ünsal, Martin Grambone, Huaiqi Huang — August 2013 - July 2017

Amputation of a hand or limb is a catastrophic event resulting in significant disability with major consequences for amputees in terms of daily activities and quality of life. Although functional myoelectric prostheses are available today, due to a lack of sensory function in the prostheses, there is not yet a solution for restoration of a natural sense of touch for persons using prosthetic limbs.

The goal of the WiseSkin project is to develop a tactile prosthesis that allows an amputee to feel pressure (in the future also - shear and temperature). This may significantly improve the quality of life for amputees. WiseSkin is a project sponsored by NanoTera and SNSF, which involves three main partners: CSEM, EPFL and BFH. The concept of WiseSkin is based on embedding miniature, soft-MEMS tactility sensors into a silicone based “skin”. It employs scalable, event- driven ultra-low-power wireless communication to convey the sensor data to the actuation control module enabling the sensors to be placed almost anywhere. A novel, stretchable subsystem for powering the device also serves as a waveguide for the wireless communication. Sensory feedback is based on the phantom mapping principle which is leveraged to provide natural sensation of touch by appropriate points on the amputee's residual limb using a tactile display.

At BFH, the work involves investigating sensory feedback, system design, final integration and developing a functional prototype. For the sensory feedback system, we aim for a non-invasive sensory substitution system, i.e., we apply feedback to the superficial skin via a different modality or to a different location of the body. The most commonly used sensory substitution feedbacks are: electrotactile, vibrotactile and mechanotactile as well as hybrid systems. The first approach is vibrotactile due to its small size, low power consumption and universal psychological acceptance. A testing system is being built up, involving an actuator array, the drivers for actuators and a microcontroller for generating signals to control stimulation patterns. The challenge is to find a suitable way to code the signal.

For the system design and final integration, one of the challenges is to develop a sensor fusion algorithm. Due to the limited space in the remaining stump, one-to-one mapping is not possible. Advanced signal processing allowing sensor data fusion is needed to convey precise information collected by the sensors to the actuators. Another challenge is to coordinate and integrate the whole system where the sub-systems have different data formats and power supply systems.

WiseSkin for tactile prosthetics

Intuitive and reliable representation of lung ventilation by electrical impedance tomography

Volker M. Koch, Jörn Justiz, Andreas Waldmann, Pascal Gaggero — April 2012 - December 2013

Menschen müssen aus verschiedenen Gründen künstlich beatmet werden. Zum Beispiel bei Verletzungen, Erkrankungen oder bei einer Narkose. Heute gibt es moderne Beatmungsgeräte. Aber es sterben immer noch zahlreiche Menschen während und sogar durch die künstliche Beatmung. Warum? Ist der Druck beim künstlichen Beatmen zu gross, wird das Lungengewebe überdehnt und geschädigt. Ist der Druck zu gering, kann ein Lungenabschnitt kollabieren. Beatmung mit falschem Druck kann zu einer Lungenschädigung und sogar zum Tod führen. Es wäre doch gut, wenn man während der Beatmung in die Lunge hineinsehen könnte um die Beatmungsparameter zu optimieren.

Zum hineinsehen in den Körper gibt es bereits verschiedene bildgebende Verfahren. Mittels CT kann man zum Beispiel ein Schichtbild aus dem Bereich des Brustkorbs erstellen. CT arbeitet jedoch mit ionisierender Strahlung, welche Krebs verursachen kann. Für eine kontinuierliche Überwachung von beatmeten Patienten benötigen wir eine andere Methode.

Die elektrische Impedanztomographie, oder kurz EIT, liefert ebenfalls Schichtbilder. Die Auflösung ist zwar geringer als bei CT, man erkennt aber dennoch Probleme bei der Beatmung. Da EIT keinen Krebs auslösen kann, darf diese Methode auch bei Säuglingen eingesetzt werden.

Wie funktioniert EIT nun? Elektroden werden um den Brustkorb herum auf der Haut angebracht. Dann speisen wir kleine, vollkommen ungefährliche Wechselströme ein. Wir messen die sich ergebenden Spannungen auf der Hautoberfläche.

Dann nutzen wir andere Elektroden für die Stromeinspeisung und messen erneut viele Spannungen. Mit all den Spannungs- und Stromwerten kann nun ein Schichtbild der Impedanzverteilung berechnet werden. EIT ist ein Verfahren, welches sicher und kostengünstig ist, sowie ausreichend gute Bilder von der Lungenfunktion liefert.

In Landquart wurde ein EIT-Elektrodengürtel für die künstliche Beatmung entwickelt. Aufgrund des grossen klinischen und wirtschaftlichen Potentials entstand ein neues Unternehmen. Die gemeinsamen Interessen von BFH und Swisstom haben zu einem KTI-Projekt geführt. Hauptforschungspartner war das Institut HuCE der BFH, beteiligt waren aber auch die HSR und die Carleton University.

Im KTI-Projekt haben wir an der BFH zum Beispiel ein Testsystem für EIT-Systeme entwickelt. Dabei platziert ein Roboterarm ein nicht leitfähiges Objekt in einem Tank. Dieser Tank ist mit einer physiologischen Kochsalzlösung gefüllt und mit Elektroden bestückt. Man kann nun zum Beispiel die Position des Objekts im berechneten Schichtbild mit der tatsächlichen Position vergleichen. Mit diesem Ansatz lassen sich EIT-Systeme charakterisieren.

EIT project

VoiSee - A Portable Electronic Vision Aid for Larger Pictures and Better Contrast, in Particular for Patients with Macular Degeneration

Volker M. Koch, Jonas Germann, Aymeric David Niederhauser, Markus Lempen, Jörn Justiz — March 2012 - August 2013

Age-related macular degeneration (AMD) is a disease which results in a loss of vision in the center of the visual field. Since AMD is currently untreatable, research is conducted towards a novel and portable vision aid within the scope of the VoiSee ® project. While the whole system ought to remain lightweight and small, a large field of view (FOV) and a comfortable viewing experience are crucial for the acceptance of portable vision aids. In this respect, VoiSee ® distinguishes itself from already existing yet much more restricted portable devices. Using a special display-optics combination, it yields a FOV of more than 60° and yet weighs less than 350 g. This electronic vision aid will considerably disburden the everyday life of AMD patients. Using VoiSee ® , these patients will be able to perform general tasks more independently and will be capable of reading smaller writings outside their homes (e.g. product inscriptions in supermarkets or departure boards in railway stations). The prototype developed within the frame of this project will be industrialized and marketed by a start-up company.

VoiSee project

Cardiac Simulator

Veit Schmid, Jörn Justiz — June 2011 - June 2012

The Cardiac Simulator is a demonstrator that shows the electrical activity of the heart and at the same time a tool to practice and test the programming of artificial pacemakers. It provides connectors for the output signals of the atria and the ventricles. These signals resemble those of real pacemaker electrodes. The Cardiac Simulator simultaneously responds to external electric tensions induced into it. Thus, the atrium and the ventricle can be stimulated or even defibrillated separately. LEDs indicate, which chamber of the heart currently receives a stimulus.

The graphical user interface provides a choice of different conduction disorders of the heart in order for the students to check their self-programmed pacemakers. All signals of the atrium and the ventricle as well as the entire ECG are displayed and different properties of atrium and ventricle signals can be tuned individually.

The Cardiac Simulator includes an operating device which can manually activate stimulations of the atrium and the ventricle.
The control logic of this “artificial heart” is programmed on the FPGA-chip of a NI-CompactRIO system. The output signal has a sampling frequency of 10 kHz and reacts in real-time to any stimulation signal. Optionally, different noise patterns can be added to the output signals.

In the specialization module “Biomedical Engineering” as part of the Bachelor degree program of the BFH-TI students work with the Cardiac Simulator during their respective lab course, so they can learn the most important functions of modern artificial pacemakers in a hands-on way. Starting with the simplest variants they program more and more complex algorithms up to dual-chamber on-demand pacemakers which detect and stimulate both the atrium and the ventricle.
Initial experience with students has shown that the programming of their own pacemakers provides an intensive examination and hence a thorough knowledge of the (electrical) activity and functions of the heart and the way medical technology can regulate conduction disorders.

Cardiac simulator

Development of intra-vaginal sensors to measure pelvic floor muscle activity

Damien Maurer, Volker M. Koch — January 2010 - December 2012

SUI is a particularly prevalent type of incontinence that is defined as the “complaint of involuntary urine leakage on effort or exertion, or on sneezing or coughing” [Morin, 2004]. Several theories have attempted to explain female urinary continence mechanisms emphasizing the importance of the PFM in urethral closure for maintaining continence [Morin, 2004]. Nevertheless, the lack of suitable commercially available instrumentation to assess PFM activity remains a significant diagnostic problem and prevents physiotherapists and medical doctors from understanding the pathophysiology of SUI.

In order to provide a better understanding of the mechanisms of PFM contraction and its influence on SUI, two chemically disinfectable intra-vaginal sensors were developed. These sensors allow complex measurements such as static and dynamic PFM strength in the transverse plane, rate of force development, electromyogram measurement, and threedimensional position and orientation changes of the probe during a PFM contraction. The intravaginal sensor prototype I is composed of four measurement units, each equipped with a force measurement platform based on thin film strain gauge technology and a pair of bipolar electrodes, while the intra-vaginal sensor prototype II is a lightweight version composed of only two measurement units. Orientation and position of both sensors within the vagina is tracked by a six degrees-of-freedom electromagnetic tracking device (trakSTAR, Ascension Technology Corporation).

Preclinical tests on a few healthy subjects showed that the developed probes are complementary: while the intra-vaginal sensor prototype I, in some cases, might prevent practitioners from studying functional contractions during physical activities (walking, climbing stairs, etc.), the intra-vaginal sensor prototype II provides less functionality but a better stability due to its reduced weight and improved design.

The results suggest that these investigational devices fully meet the defined user and system requirements and are ready for human use. Approvals from the competent Ethics Committee and Swissmedic are pending.

References: M. Morin et al, Neurology and Urodynamics, 23:668-674, 2004.

Intra-vaginal sensor project

Determination of the Transit Time through the Intestine of Cows with Cecum Dilatation-Dislocation

Volker M. Koch, Markus Lempen — December 2009 - April 2010

Cecal dilatation-dislocation (CDD) is a common and economically important abdominal disorder that affects mainly dairy cows. Affected animals show a reduced appetite, milk drop, colic and diminished or even lack of defecation due to constipation of the cecum with ingesta. Despite several studies, the pathogenesis of CDD little is known so far. Results from previous studies suggest that the cause of CDD is not in the cecum itself but can be in a more distal part of the colon.

In the planned study, the transit time between different sections of intestine (ileum, cecum, colon and rectum) in cows after CDD is to be measured. By comparing the intestinal transit times of various sections to the rectum in different animal groups, the area where the dysfunction, leading to CDD, occurs can be localized.

To determine the above mentioned transit times, a small, implantable capsule with built-in temperature sensor and wireless data transmitter was developed. The principle is based on the fact that the measured temperature will drop abruptly at the moment where the capsule leaves the intestine of the cow (drop from the body temperature of the cow (38.5-39.0°C) to ambient temperature). The data packets sent from the capsules are captured by a receiver in close proximity of the cow.

Design parameters of the capsule:

• Size: 22x8.6mm

• Power supply: 2x1.55 silver oxide cells

• Meas. period: 3s

• Battery lifetime: 15d

• Range:10m

Determination of the Transit Time through the Intestine of Cows

Fast Data Processing for Multi-Electrode-Arrays

Jamileh Rahbar, Volker M. Koch, Christian Dellenbach — December 2009 - December 2011

Multi Electrode Arrays (MEAs) provide a powerful interface to obtain an integrative understanding of the physiology and pathophysiology of excitable cells and are likely to find applications in drug screening studies. The computational needs to acquire and process the data over extended periods of time are huge and surpass the capabilities of commercially available conventional computer techniques. The development of dedicated hardware algorithms running in real-time on a FPGA to isolate biological signals, extract important parameters and discard the irrelevant parts of the data, solved this problem.

With current software solutions, the evaluation of an experiment recorded with a 64 channel MEA takes from hours up to days. Results of an experiment are not available before the evaluation process of the recorded data is completed. While performing an experiment with MEAs, researchers have no information on the actual state of their tissue under investigation (e.g. the activity of the cells). Thus, they are not able to react on special events or conditions. Each experiment is therefore very time consuming and for drug screening application not suitable. The acquisition and analysis process has to be optimized to enable researchers to perform their experiments in a pleasant end more efficient way.

Electrophysiological signals measured using a MEA have long periods (between two occurring action potentials) that are not of interest when studying cell activation and excitation propagation and therefore dispensable (see figure). Dedicated and solid hardware algorithms to detect action potentials, isolate biological signals, extract parameters of interest and discard the rest of the data were developed in this project. The algorithm runs for multiple channels (all the 64 electrodes) in parallel and was implemented using a semi-pipelined approach on a FPGA to meet the real time demand of the researchers. Such a dedicated system outperforms any software solution.

The new acquisition process has big advantages compared to the existing situation. Due to the fact that the signals are evaluated directly on a hardware level, the amount of data transmitted to the computer could be reduced to at least 90%. Through on-line extraction of physiological parameters a real-time visualization has become possible. Therewith, researchers can implement different feedback loops which open the door to a wide range of novel, unprecedented experiments.

The newly developed system not only facilitates the MEA experiments, but also allows the acquisition and analysis of the signals is in real-time. Existing experiments can be performed faster and cheaper. Furthermore can the system now be used not only for basic physiological experiments, but also for drug screening. Novel MEAs tend to have a higher amount of electrodes as well as a higher spatial resolution. The implementation of the high speed detection and extraction algorithms on a FPGA, allows an easy scalability to handle MEAs with several times more electrodes. Sole limit is the data transfer capacity towards the computer.

Fast Data Processing for Multi-Electrode-Arrays

A physiologic robot reactor system to simulate in vivo conditions

Veit Schmid, Jörn Justiz — January 2008 - January 2011

Introduction
Growth of cartilage tissue crucially depends on the interaction of cells with scaffold materials, and environmental conditions such as biochemical factors and mechanical triggers [1-3]. Two main shortcomings have been identified in existing bioreactor systems: (i) the mechanical stimulation units do not operate within a physiological stress range and they are limited in the applicable motion pattern; (ii) most systems lack an ambient control and therefore no hypoxic environment is generated as encountered in synovial joints. We have addressed these shortcomings by designing a fully autonomic modular Physiological Robot Reactor System.

Methods
We have engineered a reactor system that comprises a mechanical stimulation unit (MSU), an automatic sample changer (ASC), and an environmental control box (ECB). The MSU is designed with three linear (orthogonal axes) and one rotational degree of freedom (around z-axis; a rotational component around y-axis is pending). The load generated by the MSU is transferred via an exchangeable plunger on a sample tissue placed in a sample holder. Highly accurate force- feedback and motion systems are controlled by ultra- fast Field Programmable Gate Array (FPGA) and real- time components which continuously monitor all system parameters. The ASC is designed as a carrousel providing space for 24 sample holders, which allows for individual piloting of the samples with their own stimulation pattern. The ASC and the MSU are integrated in the ECB in which humidity, temperature, gas composition (O2, CO2), and pressure are actively controlled. In addition, an automated media exchange is also implemented in the system, which enables a prolonged uninterrupted cultivation of sample tissues.

Results
The complex physiological motion and load pattern of a knee joint [4,5] were closely traced by the precisely controlled robot axes. A large range of loading forces of from less than 1 N up to more than 300 N in longitudinal and 100 N in lateral direction were achieved, closely matching the physiological forces encountered in the knee. The precise and fast motion controls provide a position accuracy of about 10 mm. Within the ECB, the climate is accurately controlled and maintained (deviations of less than 0.2% and 0.2°C from given gas concentrations and temperature, respectively).

Conclusions
The PRRS has the potential to be a convenient and flexible tool for screening and evaluation of cartilage repair strategies in vitro performed under the harsh conditions encountered in vivo within synovial joints. The PPRS is designed modularly, thus it is a very flexible system that may also be used for the stimulation of other sample tissues than cartilage, with much different mechanical and environmental parameters.

A physiologic robot reactor system to simulate in vivo conditions