MEDICAL INTERNET OF THINGS (MIOT) & IMBEDDED INTELLIGENCE IN HEALTHCARE - DR. ABDELBASET KHALAF
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Medical Internet of Things (MIoT) & Imbedded Intelligence in Healthcare Dr. Abdelbaset Khalaf khalafb@tut.ac.za 3rd GFMD-WHO
Medical Internet of Things (MIoT) rate • Integra)on of medical devices in a network connec)on • Network can be managed from the web • Provide informa)on in real )me • Communica)on: person to person (P2P) & machine to machine (M2M) • Allow interac)on between health professionals & pa)ents • MIoT can be seen from 3 paradigms: Ø Internet-oriented middleware Ø Things sensors oriented Ø Knowledge-oriented seman)cs
Understand Internet: IP Addressing IP Addresses connect the Internet Number / Address 5 RIRs Internet Protocol LIR / ISPs Number / Address End Users
6 5 4 3 2 1 0 IPv4 Address Space Issued IPv4 Address Space Issued by year : RIRs to Customers Fixed length, 32 bit scheme, more than (Jan. 1999 – Dec. 2008) 4 billion (232) addresses Source: INR Status Report (NRO, As of 31 Dec. 2008) by H Zhao ITU
IP Next Generation Protocol IPv6 Addresses 2128 = 3.40282 x 1038 IPv6 Greatly expanded address space More attractive for (2128) future Internet applications compared to IPv4 Potential socio-economic Multi Access: beneZits for Enhanced life mobility ubiquity of the Internet IPv4: Fixed length, 32 bit scheme, more than 4 billion (232) addresses
IPv6 Deployment: Essential for wireless Internet Emergence of mobiles as platform for wireless Internet access especially in developing countries will put more pressure on the IP address space Require a larger IP address space to enable wireless networking & mobility IPv6 protocol provides the availability & extensibility of IP addresses : Large-scale sensor networks, IP Security, Mobile IPv6, IP-based Mul)media IPv6 is emerging as the preferred plaJorm and is a core component of the wireless Internet architecture (3G & Beyond 3G) Internet is now a critical global infrastructure for socio-economic development and growing faster in developing countries
Embedded Intelligence in Healthcare Insights / Proximity Act ion Low Inte llige nce latency Technical De vice Se nsor Integration Quality of Dat a Experience Virtualization Product Improve Faster time to market monitoring care plans New markets Business Embedded Healthcare Applications Transformation Revenue generation Intelligence Disease Helping detection doctor A Cloud Advance d Analyt ics / Industry Collaboration Pat t e rn Re cognit ion/ Standards Classifications/ IoT Applicat ion / Network Opt imizat ion Dat a St ore 17/05/01 TUT/FSATI 7 2017
Embedded Intelligence in M- IoT • Growth at a high rate exceeding 7% • Estimated Revenues by 2020 $2.2 trillion • Healthcare is one of the Leading industries What is it? How does it help? Embedded int elligence is the ability of a - Monitor he alth and us age of products product, proces s or s e rvice to monitor its to e xte nd t he ir performance and - Ope rational pe rformance , lifetime - us age load, - Improve marke t appe al and - e nvironme nt acce ptance of products - The ability for a s e rvice, s ystem or Goals: product to be us e d by age ing and people - e nhance pe rformance and lifetime , with special needs. - incre ase quality and - Address skills shortages in limited resources - improve cus tome r s atisfaction. - Enabling ne w re venue opportunities 17/05/01 TUT/FSATI 8 2017
Artificial Intelligence (AI) on the Edge Supported by Fog/Cloud Edge Fog Cloud Data in Mot ion Hist oric/ Pre dict ive Analytics More comput ing Healthcare Devices & Systems More int e ract ion and re s pons e Device t o device communicat ion 17/05/01 TUT/FSATI 2017 9 10 B Rapolu
Putting It All Together 17/05/01 TUT/FSATI 10 11 2017
Scenario: Intelligent MRI Machines Onboard Sensors Imaging System - Captures temperature at various - MR imaging controls positions on MR Machine q Scan Dat a q Syst em Patient Data - Body part for scan Failure Logs - Weight of patient Data Failure Event Pattern Magnetic Field Control Matching - Manual setting of magnetic Acquisition System Algorithm field required for scan as prescribed by the doctor Predictive Analytics Platform Other data sources q Equipment hist ory q Maint enance records q Environment al dat a q Expected failures q Maintenance Schedules System Health Dashboard Predic)ve Asset Optimisation Models q Predictive Asset Maintenance Maintenance 17/05/01 TUT/FSATI 11 2017
Example: Architecture of System 17/05/01 TUT/FSATI 12 2017
Success Requirements: Are we ready? Watch the outcome of Horizon 2020:eHealth workforce development • The core of any healthcare system is its workforce • Healthcare system requires a robust supply of highly skilled professionals • And they must be digitally skilled in eHealth Ø The future state of healthcare depends on workforce with eHealth skills Ø How can we address workforce shortage and the lack of access to skills/competencies in eHealth/health IT? v We need to map and quan)fy needs & supply, demands & trends for skills & competencies for all eHealth actors v The way forward: gap analysis, case studies and stakeholders engagement to form bigger picture of eHealth workforce v Development of eHealth/Health IT courses/curricula
Current developments and research domains • Body Sensors Network BSN applica)ons • Energy-efficiency for BSNs • Security and privacy for BSNs • BSN system architecture • Interference mi)ga)on in BSNs • Systems enabling pa)ent self-monitoring and assessment • Hardware for BSNs • BSNs with Cloud Compu)ng Capabili)es • BSNs for eHealth and ac)vity monitoring/biomonitoring • BSNs and wearables • Brain-2-Brain Communica)on • BSNs and the Internet of ThingsBrain-2-Brain communica)on • Expert systems for illness diagnosis in limited resources countries
You can also read