Skip to main content

International Network for Pattern Recognition of Tumours using Magnetic Resonance

Objective

We aim to facilitate the use of magnetic resonance spectroscopy (MRS) for improved diagnosis and therapy of patients with brain tumours brain tumours. Spectra, easily acquired alongside commonplace MR imaging (MRI) procedures, uniquely delineate biochemistry of human tissue in situ. Although MRS gives significantly improved brain tumour categorisation, it is not widely used, partly because radiologists have difficulty in interpreting spectral data. We therefore aim to develop a user-friendly computer program for spectral classification.

Systems development will be informed by:
(a) a large "training set" of data contributed by members of the consortium and
(b) new spectra acquired under agreed protocols.

Automated pattern recognition techniques will be developed for tumour classification together with an intuitive interface. The program will be freely available and will promote wider use of MRS, hence reducing the need for distressing and dangerous brain biopsies.

Objectives:
We will develop and test an innovative medical system for diagnosis and grading the malignancy of brain tumours and planning therapy, without a highly unpleasant brain biopsy. This Pattern Recognition program and its innovative, user friendly Graphical User Interface will accept MR spectra from advanced MRI instruments, widely available in Europe. We will also develop a large database of standardised brain tumour spectra and make it available to specialists in European Hospitals.

This project will:
(1) enable radiologists to categorise brain tumours using MRS;
(2) aid planning of treatment and therapy
(3) alleviate patient distress;
(4) facilitate the uptake of MRS by clinicians;
(5) consolidate MRS as a viable alternative to brain biopsy.

Programme Relevance: an innovative decision support system for neuroradiologists that will improve the quality of life of many citizens with brain tumours.

Work description:
The main tasks in the development of this system will be to:
1. Acquire a much larger database of compatible MR data and associated clinical data than any currently available in order to:
Develop robust methods to process and classify new spectral data according to type and grade of a tumour with statistical measure of confidence
-Develop a graphical user-interface (GUI), which will allow the user to select and display images and spectra, and also provide a flexible range of representation of the classifier output
-Combine 2 and 3 to produce a user-friendly program for clinical use.

We will install a prototype of the system with a simple point-and click GUI at the collaborating hospitals early in the project. This will act as the focus for the work and stimulate feedback from the clinicians at an early stage. It will be tested developed as the project progresses and more data are acquired, and will be changed according to user requirements.

The first prototype GUI will consist of three separate "windows":
1. The classification window will present each spectrum as a point in a multidimensional space so that the user can compare an individual spectrum of unknown class with others for which the classification is already known. Confidence parameters from the classifier will also be displayed as a further aid to clinical decision-making.
2. The image window, which will have two main purposes:
a) To show the regions in the brain (voxels) from which the spectra are acquired and
b) To present a "nosologic image" of the brain showing the classification of the different tissues within and surrounding the tumour.
3. The spectrum window. This will display individual spectra in standard format (x-y plots) together with mean and standard deviation, as well as spectra of different tumour types. Again this will be simple at first, but will have potential for further development, for example to give biochemical interpretation of selected peaks.

Milestones:
Month 3: An organised collection of spectra and associated clinical data, suitable for use in the first prototype classifier system
Month 18: A fully operational database management system and prototype classifier system
Month 24: A Prototype Classifier System
Month 36: The Decision Support Tool, and the Definitive Database.

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

UNIVERSITAT AUTONOMA DE BARCELONA
Address
Campus Universitari S/n
08290 Bellaterra (Cerdanyola Del Valles)
Spain

Participants (9)

CENTRE DIAGNOSTIC PEDRALBES, S.A.
Spain
Address
Calle Londres 2
08029 Barcelona
INSTITUT DE DIAGNOSTIC PER LA IMATGE
Spain
Address
Autovia De Castelldefels Km 2,7
08907 Hospitalet De Llobregat
PERCEPTION RAISONNEMENT ACTION EN MEDECINE
France
Address
4, Avenue De L'obiou
38700 La Tronche
SAINT GEORGE'S HOSPITAL MEDICAL SCHOOL
United Kingdom
Address
Cranmer Terrace
SW17 0RE London
SIEMENS AKTIENGESELLSCHAFT
Germany
Address
Wittelsbacherplatz 2
80333 Muenchen
STICHTING KATHOLIEKE UNIVERSITEIT
Netherlands
Address
Geert Grooteplein-noord 9
6525 EZ Nijmegen
STICHTING KATHOLIEKE UNIVERSITEIT
Netherlands
Address
Geert Grooteplein-noord 9
6525 EZ Nijmegen
THE UNIVERSITY OF SUSSEX
United Kingdom
Address
Sussex House Falmer
BN1 9RH Falmer, Brighton, East Sussex
UNIVERSITE JOSEPH FOURIER GRENOBLE 1
France
Address
621 Avenue Centrale - Domaine Universitaire
38400 Saint Martin D'heres