infso-ri-508833 enabling grids for e-science a grid approach to distributed image analysis for...

19
INFSO-RI-508833 Enabling Grids for E-sciencE www.eu-egee.org A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia Torterolo University of Genoa / DIST EGEE User Forum CERN, 01-03.03.2006

Upload: meghan-nelson

Post on 21-Jan-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

INFSO-RI-508833

Enabling Grids for E-sciencE

www.eu-egee.org

A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease

Livia Torterolo

University of Genoa / DIST

EGEE User Forum

CERN, 01-03.03.2006

Page 2: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 2

Enabling Grids for E-sciencE

INFSO-RI-508833

Agenda

• Introduction to SPM analysis and to development of a web-based SPM service

• GRID implementation of SPM service

• Current Status and Plans

Page 3: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 3

Enabling Grids for E-sciencE

INFSO-RI-508833

Partners

• The first part of the work has been done during the Italian Neuroinformatics OCSE Node project in collaboration with

– San Raffaele Hospital of Milano– University of Milano – Bicocca

• GRID implementation has been carried out by Bio-Lab at DIST, University of Genoa, during the Grid.it Italian project

• The application is a quite stable work in progress and served as a demostrator for the Grid.it project but further developments and extensions are in progress still inside the GILDA testbed

Page 4: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 4

Enabling Grids for E-sciencE

INFSO-RI-508833

Introduction to Alzheimer Disease

• AD is the most common form of dementia, accounting from 50% to 70% of all dementias in elderly people (around 3 million people in EU25)

• Clinically, AD is characterized by a progressive loss of cognitive abilities

• Memory loss is typically the earliest sign of AD

Page 5: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 5

Enabling Grids for E-sciencE

INFSO-RI-508833

Analysis of PET-SPECT images

Page 6: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 6

Enabling Grids for E-sciencE

INFSO-RI-508833

Results of a SPM analysis on a PET study of glucose metabolism in a patient with dementia. Ipometabolic pattern in the frontal cortex: design matrix (top right), statistically significant clusters on a glass brain in three orthogonal planes (top left) and on a 3D brain rendering (bottom)

SPM Results

Page 7: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 7

Enabling Grids for E-sciencE

INFSO-RI-508833

Why a portal?

• The remote access to SPM analysis could provide doctors from peripheral hospitals with an invaluable tool to run the analysis remotely only using a standard browser web

• No particular HW resources or computer knowledge are required

• In order to avoid errors during the analysis, only selected users should access and use the service

http://www.neuroinf.ithttp://www.neuroinf.it

http://www.neuroinf.ithttp://www.neuroinf.it

Page 8: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 8

Enabling Grids for E-sciencE

INFSO-RI-508833

SPM: Why GRID?

• A large set of images of normal patients is required to be used for comparison

• PET and SPECT studies on normal subjects are very rare

• Images of patients are covered by privacy and security issues and for this reason they cannot be freely moved on the net or published by the centre that made the analysis

Page 9: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 9

Enabling Grids for E-sciencE

INFSO-RI-508833

Contribution to GILDA

• LCG Grid Environment v. 2.6.0.

6 machines dedicated to LCG NODE:

1 CE, 1 SE, 3 WNs, 1 UI, 1 LCG install server (12 CPUs, 7Giga (RAM), ≈ 350Giga of storage)

Page 10: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 10

Enabling Grids for E-sciencE

INFSO-RI-508833

• Data Management and File Access tools:

to access data from User Interface using Logical File Names (LFN)

• LCG File Catalog (LFC):

to register data into the catalog

• AMGA Metadata Catalog:

to add related metadata to data

LCG tools used

Page 11: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 11

Enabling Grids for E-sciencE

INFSO-RI-508833

Data Management and File Access

Page 12: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 12

Enabling Grids for E-sciencE

INFSO-RI-508833

LCG File Catalog (LFC)

Several functionalities:

• Use of lcg_utils• Use of GFAL calls• Use of GSI certification

• Access to grid file in SEs from “anywhere”• Several replicas of files in different sites• Copy of data from/to local file system to GRID

Page 13: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 13

Enabling Grids for E-sciencE

INFSO-RI-508833

AMGA: ARDA Metadata Grid Application

Side-by-Side a File Catalogue (LFC): File Metadata

Access control to resources on the Grid is done via VOMS

Different front-end:– mdcli & mdclient and C++ API– Java Client API and command line

mdjavaclient.sh &– mdjavacli.sh (also under Windows !!)– Python Client API

Strong security requirements:– patient data is sensitive– metadata access must be

restricted to authorized users

Page 14: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 14

Enabling Grids for E-sciencE

INFSO-RI-508833

GRID-Design of the application

With reference to different LCG tools mentioned above, application code has been modified in the following way:

– Registration and storage of data files (PET/SPECT images) on SEs available using lcg_utils interacting with LFC and AMGA.

– Development of a C program with GFAL C API in order to access distributed images using their LFNs and to extract some information necessary to SPM analysis without copying them locally.

– Job Submission: creation of a JDL file to submit the executable (and not the images) with GFAL call to the GRID.

– Statistical Analysis: running of SPM analysis from results obtained from job submission. Statistical analysis is performed outside GRID environment.

Page 15: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 15

Enabling Grids for E-sciencE

INFSO-RI-508833

Final GRID structure

Page 16: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 16

Enabling Grids for E-sciencE

INFSO-RI-508833

Application Running: Setting Parameters

Setting SPM parameters Selection of normal subjects

QUERIES to AMGA !

Page 17: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 17

Enabling Grids for E-sciencE

INFSO-RI-508833

Application running: JOB SUBMISSION

SPM Analysis RESULTS

Page 18: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 18

Enabling Grids for E-sciencE

INFSO-RI-508833

Issues & Problems

• Pros:– Data distribution over different SEs accessed with GFAL APIs.– Integration of a Grid application into an existing environment with

hard constraints and several integration problems to solve– Application behaviour to the user unchanged

• Cons:– GFAL API not well documented & Compatibility issues and

conflicts on libraries (but good support from GILDA staff)

– Too much components (Zope + Plone, Python, shell scripts, Matlab, C/C++ compiled code, …& Grid!)

– Software maintenance is hard

Page 19: INFSO-RI-508833 Enabling Grids for E-sciencE  A Grid Approach to Distributed Image Analysis for Early Diagnosis of Alzheimer Disease Livia

EGEE User Forum, CERN, 01-03.03.2006 19

Enabling Grids for E-sciencE

INFSO-RI-508833

Current Status and Plans

• Application– pre-processing parallelization (further functions to port on Grid)

• Interface– move to Genius 3.1

• Overall framework– improve performance and scalability– A Grid environment devoted to the analysis of microarrays, derived

from this project, will be tested  in the frame of European Project BioinfoGRID http://www.itb.cnr.it/bioinfogrid (Bioinformatics Application for Life Science)

– A bioinformatics data storage and analysis platform with metadata support will be developed under Italian MIUR-FIRB project LITBIO www.litbio.org (Laboratory for Interdisciplinary Technologies in Bioinformatics)

Thanks for your attention!