Anzeige

An Event of

Powered by

To use this page, you need JavaScript enabled

14. November 2019
ICM Munich

electronica Medical Electronics Conference

Complete Program

Thursday, 14. November 2019

09:00 - 10:30
Welcome & Keynote Session Keynote
09:00 - 09:30
Welcome and Registration
09:30 - 10:15
Keynote: Monitoring to Improve Health & Performance Prof. Dr. Billy Sperlich, Julius-Maximilians-Universität Würzburg  
24h-Monitoring to Improve Health & Performance – (Why Aren’t We) on the Way to Precision Health & Training!?
10:15 - 10:30
Short Break to Select Tracks

Parallel Sessions

10:30 - 11:30
Session 1 Medical Device Regulation
10:30 - 11:00
Working Title: EU Medical Device Regulation Gesa Kniebühler, TÜV SÜD
11:00 - 11:30
Artificial Intelligence and MDR – Verification possible? Thorsten Langenhan, AVQ  
Artificial Intelligence (AI) has come into application maturity with the application of deep learning and machine learning, in which it is also interesting for use in safety-related areas. As an example, the feature recognition in pictures or videos may be mentioned here and, for example, the further processing of the found features in the automated diagnosis. The physician's support in making diagnostic decisions is mentioned here as an end-to-end example of a possible area of safety application of the AI. Medical devices are subject to the new MDR. As a general standard for the technical implementation of MDR and the risk management for electrical / electronic controls, the standards IEC 60601 for hardware and IEC 62304 for the software apply. If, as shown in the example, diagnoses such as the detection of a tumor in a recorded X-ray image are carried out with learning algorithms, then this diagnosis must be checked for correctness that the accepted dangerous error rates can be fulfilled by the regulations of the Regulatory Affairs. This means, on the one hand, that the results from AI algorithms have to be checked "deterministically". For this purpose, usually safety analyzes are necessary, which show with what probability the information content of the signal can be verified as correct. Thus, statistical and also systematic checks are necessary to ensure the truth content of the signal. Approach: In the case of a supervised learning algorithm, for example, the safety-critical features can be stored in the "supervisor". A big strength of AI compared to conventional graphics tools is that context also has an influence. Especially under the aspect of ”Adversial Effects "AI evaluations of images can achieve a dramatically different evaluation than the AI was previously trained to. In practice, this spoofing can lead to significant misinterpretations, which may possibly mislead the physician. The expense of proving the correctness of the signal can be considerably high, since the verification extends over all application areas and over the entire life cycle of the product. In the example of automated diagnosis, all special shapes of tumors and other circumstances occurring in reality should be considered. Tests can certainly help, but will only be able to cover a subarea. However, because the range of tumor shapes, taking into account other shadowing in the image as well as the status of an AI method, has a very high variety of states, proof of functional safety becomes analytically impossible and statistically so complex that one can also speak of impossibility to achieve required confidence levels. So what helps? On the one hand, the diagnosis can and must run on further indicators, the automated interpretation of e.g. An X-ray alone can lead to significant misdiagnosis. On the other hand, the image interpretation remains based on pure deterministic algorithms and tools, which are also constantly evolving, which have verifiable informative value.
10:30 - 11:30
Session 2 Connected Healthcare
10:30 - 11:00
Health-IT: One Step Further in Making our Hospitals Digital Dr. Karsten Königstein, Sinfosy  
There are numerous obvious benefits of digital technologies for healthcare providers: ✓Less paperwork, reducing of time costs – up to 60% ✓Localization and tracking of the goods and people in the real-time ✓Using outcomes information to improve decision-making But what are the main challenges healthcare providers meet? First of all, this is the problem in patient observation: the traditional way with handwriting notes is time-consuming and could lead to errors. The second problem is tracking of medication and equipment in hospitals. Where is my infusion / sonogram device at the moment? What quantity of a particular drug is still available? And further examples of use... SINTRA is the key! ✓Real-time localization of patients and medication / equipment with help of active tags and standard hardware (active and passive) ✓Simple inventory through real-time stocks ✓Tracking of movements in the hours or across locations ✓Elimination of all searches How SINTRA solution work in the hospital? A master folder is marked with a number and a tag. The patient file is inserted into this folder and accompanies the patient on his way through the departments. There is no assignment of patient to U file number on the server as well as no patient data is stored! The data can be evaluated in different ways –treatment time waiting period per patient and time-dependent while linkage of the evaluation with the patient name can only be carried out by the hospital. In conclusion, the benefits of healthcare digitalization: ✓Simplifying organizational processes ✓Improvement of treatment quality and patient care ✓Reduction of paper-based processes ✓Just Plug & Play: no need in IT-Project, no special hardware required More than 75% german hospitals will start the process of digitalization by the end of 2019 All processes of tracking & localization in the hospitals will be performed via Apps ….
11:00 - 11:30
Bridging the Digital Divide: A Practical Guide to Connected Healthcare David Niewolny, Real-Time Innovations (RTI)  
Healthcare is on the verge of a technological transformation. Ubiquitous connectivity, artificial intelligence and machine learning have the potential to radically transform the healthcare industry. The key to leveraging these new advances in technology is the emergence and adoption of a connected healthcare platform. Connected healthcare expert David Niewolny will discuss the current state of the market, introduce the idea of a connected healthcare platform and discuss the resulting benefits of executing on this approach. He will also discuss the challenges of architecting, developing and implementing these interoperable, highly reliable, semi-autonomous, real-time systems. He will review the open architecture needed to create a secure, reliable and interoperable foundation for development. Specific topics that will be addressed in this session include: How to effectively address the wide range of demanding data connectivity requirements How to facilitate plug-and-play interoperability How to enable the transport and analysis of unprecedented amounts of data in real time How to provide a level of system security that keeps your network and data safe How to eliminate single points of failure to create a highly reliable, “medical grade” network
11:30 - 13:00
Keynote Session & Lunch Break Living with a Bionic Hand
11:30 - 12:00
Keynote: Living with a Bionic Hand Michel Fornasier, aka Bionicman
12:00 - 13:00
Lunch Break and Networking

Parallel Sessions

13:00 - 14:00
Session 3 Chronic Diseases
13:00 - 13:30
A Novel Wearable Device for Monitoring Lung Function from Tidal Breathing Prof. Dr. Wieland Voigt, Nanovation-GS  
Chronic obstructive pulmonary disease (COPD) is a global healthcare burden with >250M patients world-wide. It is the 3rd leading cause of death and a major cost driver, mainly due to repeated hospital admissions and readmission of patients due to clinical deterioration. Detection of early signs of deterioration is believed to improve this situation. However, as of now no validated technology to allow self-monitoring of patients outside the hospital is on the market. NanoVation is developing Respiratory Monitoring Technologies and Wearable Medical Devices, utilizing novel nano-based sensor technology licensed from the Technion. These unique sensors provide high-resolution measurement of the respiratory waveform, enabling for the first time a completely seamless and passive monitoring of lung function. We think that the technology will be able to address clinically unmet needs in COPD and therefore NanoVation is targeting early detection of COPD exacerbation and reduction of the associated financial burden (~$14B annually in the US alone). NanoVation’s product, the SenseGuard™, is an innovative wireless wearable device that enables continuous and non-invasive monitoring of critical respiratory parameters derived from patients' tidal (normal) breathing waveform, cancelling the need for forced expirations. The current device operates as “stand-alone”, where the data is logged, analyzed and displayed on a tablet. The product’s first generation is pending CE marking approval, expected during 2019. A cloud-based solution for data-analysis and storage including a web-enabled interface will be developed during 2019-2020. As validated in two clinical studies, the SenseGuard™ provides clinically accurate respiratory rate measurements as good as the expensive and complex gold standard Capnography. Furthermore, SenseGuard™ demonstrates an unmatched ability to identify even the mildest changes in inhalation/exhalation duration and ratio, which reflect changes in a patient’s respiratory condition, airway obstruction and restriction. Bench testing with breath simulator proved that NanoVation’s sensor time resolution is of at least 0.1 seconds, which is more than enough to identify meaningful changes in these parameters. Preliminary clinical results already showed that the E/I ratio of 7 hospitalized patients without any respiratory conditions were between 1.63+/-0.34 to 2.51+/-0.33, while 2 Asthma patients showed higher values of 3.38+/-0.62 and 4.07+/-0.50 (Asthma as background condition). Moreover, a COPD patient that was admitted with an acute COPD exacerbation, presented an even higher value of 5.59+/-0.85. The current data suggest that the SenseGuard™ could distinguish regular from abnormal respiratory conditions and warrant further clinical evaluation. Next clinical studies will aim to validate the ability to provide accurate assessment of a changes in the respiratory condition (improvement or deterioration) of COPD patients.
13:30 - 14:00
TBD Peter Ferguson, Arm  
According to the World Health Organization, more than 400 million people live with diabetes globally. The disease, if left untreated, can lead to serious health complications such as blindness, amputation, or even death. This presentation begins by outlining the significant current events and industry trends taking place in the diabetic care space to set the stage for the balance of the presentation. Topics that will then be discussed include the need for accurate blood glucose measurement and insulin delivery, the different methods and system architectures which can be applied, as well as the range of technology that makes it possible, bearing in mind that these devices are becoming smart, cloud connected, and most even body worn. The ‘holy grail’ of the technology – the artificial pancreas – is also discussed, showing various system design considerations and implementation options ranging from edge-node biometric monitoring and therapy delivery, to cloud connectivity. Conclusion: ON Semiconductor is improving lives through innovative medical solutions. The speaker will explore 3-5 underlying critical characteristics for devices in diabetes management. The audience will learn to think about key critical characteristics when adopting diabetes management technologies. There are many essentials to consider, including major attributes like long battery life, high accuracy/clinical performance, and small form factor, as well as other ‘soft’ key characteristics.
13:00 - 14:00
Session 4 Big Data
13:00 - 13:30
5G and Healthcare: Real World Performance Data on 5G Shows Potential in Healthcare Roeen Roashan, IHS Markit  
The talk will focus on the impact of 5G on healthcare delivery including the extension of the clinical arm. This involves telehealth and remote monitoring services, and a better coordination of care through real-time care team communications. Ultimately, the proposition is a virtualization of today’s hospital, and a mobilization of the healthcare workforce. Central to the talk will be RootMetrics IHS Markit’s recent study of the South Korean 5G deployment. This is study that includes real world insights based on our team’s 1532 m2 travel in South Korea to test several networks inside homes, on the road and in commercial buildings. The study includes speed data and reliability data. Beyond a technological review, behavioral insights will also be showcased to depict the adoption pattern of healthcare sectors. Key to technology adoption is to understand cultural developments in local healthcare markets.
13:30 - 14:00
Challenge Big Data in Biomedicine: Long-term High-Resolution Impedance Manometry (LTHRIM) of the Esophagus Dr. Alissa Jell, Klinikum rechts der Isar der TUM  
High-resolution manometry of the esophagus is the gold standard in the diagnosis of esophageal motility disorders [1]. As the measurements take place in clinic for a very limited period of time only little is known about circadian changes and there is no recording of intermittent symptoms in esophageal motility. However, disorders of esophageal motility can lead to severe symptoms even if they occur (only) sporadically. For this purpose, a combined impedance-manometry probe with 36 pressure sensors and 17 impedance fields (Unisensor AG, Swiss) with a mobile data logger (MALT, Standard Instruments, Karlsruhe) were used. With a sampling rate of 50 hertz and a prolonged examination time of a full of 24 hours, very large amounts of data (8GB, respectively about 212 mio. entries per patient) are quickly produced. In the case of manual evaluation, an experienced physician needs three working days to evaluate a long-term data recording containing about 800 to 1500 acts of swallowing as well as a various number of esophageal motility phenomena (e.g. diffuse esophageal spasm, tertiary peristalsis, …). Therefore, we implemented a combined matlab and python scripting for extracting every pressure event of the whole examination. In the first step an automated swallowing detection algorithm based on applying Bayes’ theorem [2] combined with a loss-and-risk-analysis was developed. After cross-validation and minimizing the cv-risk we were able to establish a risk of 0,11 to our dataset compared to the risk of naive bayes with 0,61. A typical long-term measurement generates around 2000 images which were standardized extracted in 640x480 pixel2 images (fig. 1) and used for clustering. Orange, a python library, commonly used for image embedding, was modified and trained with characteristic volunteer data sets as well as defined diseases (e.g. achalasia, Zenker’s diverticula, fig. 2). We validated the most common imaging neuronal networks (Inception v3, VGG16, VGG 19, Squeeze Net, Painters, DeepLoc) for inter- and intra-observer accordance. 1. Roman S, Pandolfino J, Mion F. High-resolution manometry: a new gold standard to diagnose esophageal dysmotility? Gastroenterologie clinique et biologique 2009; 33: 1061-1067 2. Rish, I. An empirical study of the naive Bayes classifier. IJCAI Workshop on Empirical Methods in AI. 2001
14:00 - 15:00
Keynote Session & Coffee Break Predictive Healthcare
14:00 - 14:30
Keynote: Predictive Healthcare Markus Bönig, Vitabook
14:30 - 15:00
Coffee Break and Networking

Parallel Sessions

15:00 - 16:30
Session 5 Wearables
15:00 - 15:30
Wearables And AI In Healthcare: Opportunities And Threats Markus Vogt, EBV Elektronik  
The presentation gives an overview of the amazing opportunities with Artificial Intelligence in the Healthcare market. But there are also threads in using AI and connected devices, especially with data security. Another topic in the presentation are the pro and cons of using wearables in a telemedicine system.
15:30 - 16:00
Tools not Toys – Regulated Medical Devices Bundled with Phones Prof. Dr. Chris Elliott, Leman Micro Devices  
High-quality cameras, GPS receivers and music players use hardware that is now effectively free, bundled with the phone, but all were expensive extras a few years ago. Because we all have them, they have created massive new markets for apps that use them and have devastated the makers of stand-alone or add-on devices. The next revolution will be medical devices bundled with the phone; not toys that play with fitness but regulated medical devices, with clinical accuracy and diagnostic application. Instead of measuring what is easy, they will measure what is useful. The challenge is to make devices that are small enough and cheap enough to go into a phone and which make accurate clinical measurements and satisfy the constraints of the European Medical Devices Regulation, FDA’s clearance, China FDA’s procedures and all of the others. This paper will illustrate that challenge by presenting a device that is already in quantity production and which is being evaluated by several phone makers. From there, it will look to the future for other such devices and the medical benefits that they will bring. As well as the direct benefits of early detection and timely monitoring of disease, their data will transform diagnostic AI and telemedicine.
16:00 - 16:30
Technology for Smart Medical Wearables Enables Better Health Monitoring Andrew Baker, Maxim Integrated  
Worldwide Healthcare costs have reached $9 Trillion dollars worldwide and continue to grow at a rate higher than inflation, causing alarm to governments, healthcare providers as well as to individuals. There is a real need to find a solution to this problem. Using personalized wearable healthcare solutions is one of the ways how a more efficient delivery of care is being developed, enabling better long- and short-term outcomes. Wearable healthcare technologies are giving us real-time insights into our well-being, from heart rate and blood-oxygen levels to body temperature, sleep patterns, and much more. With this data, we can more closely monitor chronic conditions, be more proactive about preventive care, and perhaps even detect issues of concern before they happen. For design engineers, the push is on to utilize sensors, power-management ICs, and other technologies to create new types of devices and uncover new use cases. It’s all about delivering continuous monitoring of various health parameters. This is causing an increased demand for accurate, small, and low-power wearable devices. Technologies must be able to perform well in a variety of form factors. For example, optical pulse oximeter/heart rate sensors and ECG & BioZ AFE deliver accurate, continuous monitoring in compact, low power solutions. The presentation will explore 3-5 underlying technology characteristics for medical wearable end applications and outline how technology which incorporates the four wearable essentials (low battery life, high accuracy, small form factor, and clinical performance) is enabling predictive and more accurate health monitoring.
15:00 - 16:30
Session 6 Lower the Risks
15:00 - 15:30
Intuitive Orchestration of Context-Sensitive Surgical Assistance Systems Lukas Bernhard, Klinikum rechts der Isar der TUM  
Modern operating theatres hold a great variety of medical devices and assistance systems which offer increasing functionality and greater intraoperative flexibility. However, the complexity of operation and the amount of potential sources of error inevitably increase. To counter this, surgical workflow recognition systems aim at apprehending the current situation in the operation room autonomously. This provides the necessary input for context-depended assistance systems which support the surgical team with the right action at the right time. While a multitude of approaches for workflow recognition have been proposed, few work focusses on the context-sensitive orchestration of surgical assistance systems. In this work, we present a framework for intuitive configuration of workflow assistance behavior directly by the surgeon. Our proposed architecture consists of a recognition engine for detecting the current intraoperative situation and an assistance engine for processing this information and adapting the state of the technical infrastructure to fit the current needs of the surgical team. The exact behavior of the assistance engine can be specified preoperatively by defining rules that map a specific context to a desired supportive action. We argue that the definition of these rules should be handled by the surgeon personally, since he or she possesses the necessary domain knowledge and is able to model the behavior according to his or her individual preferences. To assist the surgeon with this task, we developed an intuitive graphical user interface for defining assistance rules, while not expecting any technical knowledge from the user. We evaluated our software by conducting a preliminary study with 6 experienced surgeons and 5 PhD students. We used a multi-aspect rating system to evaluate usability and user experience aspects. Overall response of the participants was positive throughout. After having received a short introduction to the user interface, the participants were able to confidently model the workflow behavior prescribed by the study protocol. We subsequently asked the participants to rate our software with regard to several usability and user experience aspects within an interval ranging from -3 (worst) to 3 (best). All aspects received ratings within the positive range (perspicuity: 1.76; efficiency: 1.77; dependability: 1.82; originality: 1.46; stimulation: 1.41; attractiveness: 1.76). Results show that a flexible definition of workflow assistance behavior by the surgeon is feasible and desirable. However, it is necessary to conduct a broader study to confirm our results with a greater number of participants and find new clues for optimizing the rule definition language and its visual representation. One central aspect is to increase the expressiveness of the rule definition language while maintaining its perspicuity, such that the resulting workflow assistance behavior is still perfectly transparent for the user.
15:30 - 16:00
500.000 Recalled Pacemakers, 2 Billion $ Stock Value Loss Tobias Zillner, Alpha Strike Labs  
During an independent security assessment of several pacemaker vendors multiple lethal and highly critical vulnerabilities were found. Based on previous experience with one specific vendor a new way of monetising vulnerabilities has been chosen. After going public a huge discussion on vulnerability disclosure ethics and responsibilities began. The stock value of the affected vendor dropped by 2 billion dollar just in one single day. The security researchers got discredited and a huge lawsuit was started. After a year of mutual accusations and denial more than 500.000 pacemakers got recalled. This talk will provide insights into pacemaker security and share first-hand experience gathered during this project. A special focus will also be on ethical vulnerability disclosure and lessons learned for future security research.
16:00 - 16:30
Security Testing Medical Apps Wilfried Kirsch, softScheck; Prof. Dr. Hartmut Pohl, softScheck  
Background/Question The number of available medical apps is constantly increasing. According to surveys, more than 50% of doctors use corresponding apps during their work. But also private people increasingly rely on medical apps. Be it for diagnosis or to always have your own medical profile with you, which can then also be presented to doctors. One current example here is a health insurance app that was nevertheless in the headlines with its security gaps, which were indeed patched/corrected quickly. When it comes to medical data, it is of the highest priority that it be secure: write- and readable only by authorized persons. The question therefore arises whether medical apps protect this data. Requests for services that indicate an illness (symptoms) must also be appropriately secured and anonymous. In addition, the availability of the stored data must be ensured! Methods Using a man-in-the-middle attack, the data sent between the app and the server was read and evaluated as far as possible. It was also tested whether the data could be manipulated undetected by the app. It was also examined where the data is stored on the smartphone. Depending on the storage location, other apps can read the sensitive medical data as well as the medical app and send it to an attacker. Outcomes The results were naturally different depending on the app developer. Few apps appear safe, because many medical data could be read and often even changed. Despite the sensitive data, which can even cost lives in the worst case scenario, many apps are still insecure. State of the art To avoid such security gaps, a holistic process is required. This not only secures the medical app, but all communication channels of the app. In order to achieve this, it is necessary to develop safely from scratch. Status of security testing the use of a security testing process with 6 methods over the entire life cycle of software development, according to ISO 27034: (1) Security Requirements Analysis, (2) Threat Modeling of the security design, (3) Conformance Testing, (4) Verification of the source code with Static Source Code Analysis, (5) Penetration Testing and (6) Dynamic Analysis Fuzzing. Each of these methods identifies different security vulnerabilities. The presentation reports on the experiences with the ISO 27034-based Security Testing Process and also deals with other evaluation criteria - in particular how far they also cover the technical level and deal with security methods for testing.
16:30 - 17:30
Closing Panel Discussion and Start-up Award
16:30 - 17:00
Closing Panel Discussion: More Technology, Better Medicine?
17:00 - 17:30
Get-Together and M+e Start-up Award Ceremony

Registration Form & Program

download

Contact

 Nina Thonesen, WEKA FACHMEDIEN

Nina Thonesen
Veranstaltungsreferentin
WEKA FACHMEDIEN

Phone: +49 89 25556 1323

WEKA-FACHMEDIEN-EVENTS

In close cooperation with recognized experts and the responsible editorial staff
of our well-known media brands we organize around 50 national and international B2B congresses, seminars and workshops each year for a defined specialist audience.

These cover topics such as embedded systems, automotive Ethernet, battery technology,
data centers, safety & security, electronic displays, Blockchain, IoT, wiring harness, AI, Smart Building, Digital Workplace, TSN, wireless technologies, cabling and much more.

Overview of all Current Events