Demonstrate that the grid can improve pre-operative planning & near- real-time surgical support by providing access to advanced simulation and image-processing services.
Build middleware on existing or developing grid technology standards to provide support for authorization, workflow, security and Quality of Service aspects.
Develop, evaluate and validate a test-bed for the GEMSS system, including its deployment in the end-user’s working environment.
Anticipate privacy, security and other legal concerns by examining and incorporating into its grid services the latest laws and EU regulations related to providing medical services over the Internet.
GEMSS Applications Medical end-users; Doctors, researchers On demand Medicine – nuclear / in vivo diagnostics Advanced image reconstruction Medical end-users; Doctors, researchers On demand Medicine – blood flow dynamics Cardio-vascular system simulation Medical end-users; Doctors, researchers On demand / distributed supercomputing Medicine – air flow dynamics Inhaled drud delivery planning Medical end-users; Doctors, researchers On demand / distributed supercomputing Medicine – Monte Carlo treatment simulation Radiotherapy planning Medical doctors, researchers On demand Medicine – intra-operative planning Neurosurgery support Medical doctors, researchers Distributed supercomputing / On demand Medicine – pre-surgical planning Maxillo-facial surgery simulation Users Class Domain Name
Maxillo-facial Surgery Simulation
Provide a virtual try-out space for the pre-operative planning of maxillo-facial surgery.
Based on image processing, meshing & HPC Finite Element simulation.
Grid scenario: client-server, doctor uses a client machine to access remote HPC services.
Modeling the distraction procedure Pre- and post surgery Courtesy Dr. Dr. Th. Hierl, University Clinic Leipzig.
An Example Implementation
Based on Web Services technologies
Client-side pluggable component framework
Protocol-independent client-side APIs for service invocation
Mean queue wait time increases and resource utilization decreases.
Schedulable only with limit percentage of ARs.
How to prevent ARs from taking start time advantages over queued jobs?
How to satisfy ARs as far as possible without sacrificing queue efficiency ?
COSY Policies for ARs …
A mandatory shortest notice time for ARs is defined
Using current mean wait time of queued jobs
Using times of mean wait time of queued jobs
Queue wait time
COSY takes the current mean wait time of queued jobs as the mandatory AR shortest notice time so that ARs cannot take start time advantages.
The mandatory AR shortest notice time is applied in COSY as 1 time of current mean wait time of queued jobs and increases to 4 linearly as the AR percentage increases from 0% (not inclusive) to 15% so that ARs can be satisfied as far as possible without increasing the queue wait time.
When the AR percentage is more than 15%, COSY will stop accepting ARs temporarily in order to guarantee a proper queue scheduling.
Advantage: adaptive to queue workload.
Disadvantage: additional calculation of statistics.