Premium
SU‐E‐I‐69: A Cloud Based Application for MRI Brain Image Processing
Author(s) -
Kagadis G,
Alexakos C,
Papadimitroulas P,
Mihailidis D,
Megalooikonomou V,
Siablis D,
Papanikolaou N,
Karnabatidis D
Publication year - 2015
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4924066
Subject(s) - cloud computing , computer science , upload , pipeline (software) , image processing , data processing , software as a service , software , database , operating system , artificial intelligence , image (mathematics) , software development
Purpose: In this study we present a cloud based application dedicated to MRI brain image processing, providing the flexibility to end users for managing and processing medical image data, dynamically allocating computational resources. Methods: The proposed application is based on the Software as a Service (SaaS) cloud model and it can be accessible by any available browser. The developed cloud takes advantage of the Infrastructure as a Service (IaaS) in order to easily scale its capabilities. Virtual Machines are used on demand of the users’ requested pipeline. The implemented techniques were incorporated on a cloud application framework consisting of a web interface where the user can upload data (NIfTI format) and initiate the processing algorithms. The proposed application is providing the following functionalities: a) a user‐based authentication and authorization mechanism, b) an application data manager which is a tool enabling the user to manage his/her uploaded image data, c) a “Data Processing Task” where the user is able to select the appropriate processing algorithm to define his/her processing pipeline on demand. Results: The proposed implementation has already been tested in several end‐users’ devices executing a pipeline for brain tumor detection in followup scans. A pipeline processing has been implemented, where the user uploads two scans of a patient (before and after treatment), they are registered, normalized and a difference map is calculated for quantitative evaluation of tumor growth or reduction. Results are stored in the cloud storage space and are available to the user anytime, at request. Conclusion: Cloud computing is one of the current biggest trends in Information Technology providing and accessing computing resources via Internet adhoc. Our goal is to offer a cloud implementation providing a way to deploy medical image processing as an application, enabling algorithm sharing without the hustle of maintaining complex infrastructures. This research has been co‐financed by the European Union (European Social Fund ESF) and Greek national funds through the Operational Program “Education and Lifelong Learning” of the National Strategic Reference Framework (NSRF) Research Funding Program: Thales. Investing in knowledge society through the European Social Fund.