Servicio de Información Comunitario sobre Investigación y Desarrollo - CORDIS

PVELab software package

PVELab is a software tool for the integrated analysis of brain PET/SPET and MRI studies. It comprise at least one module for each step needed to correct Nuclear Medicine data for partial volume effect, and software interfaces to use alternative widespread software modules. It includes interfaces so the methods can be used with data formatted in the Analyze and DICOM data format. In particular PVELab includes:

Software for loading and converting Analyze and DICOM image files, developed in collaboration by NRU, Neurobiology Research Unit, Copenhagen and IBB, Biostructure and Bioimaging Institute, Naples.

- Interactive Image Overlay developed by NRU.
- Interactive Point Selection developed by NRU.
- Interface to SPM co-registration (Statistical Parametric Mapping, Functional Imaging Laboratory, London).
- Interface to AIR registration (Automatic Image Registration, Laboratory of Neuroimaging, UCLA, Los Angeles).
- Load AIR file developed by NRU.

- QMCI segmentation developed by IBB.
- Interface to SPM segmentation (Statistical Parametric Mapping).
- Interface to BrainSeg segmentation (KIMSH, University of Kent at Canterbury).
- Load segmentation volume.

- Interface to ResliceWarp (Brain Warp, Informatics and Mathematical Modelling, DTU, Lyngby, Copenhagen).
- Interface to AIR reslicing (Automatic Image Registration).

Reslice developed by NRU Atlas (modules for automated definition of volumetric regions of interest (ROI)):
- Talairach based developed by IBB.
- MNI based developed by IBB and Inserm.
- Applyrois interface developed by NRU.

PVE correction:
Partial volume effect correction (PVEc), developed by IBB, following the methods described by Meltzer et al. [J Comput Assist Tomogr 1990;14:561-570] (M PVEc), Müller-Gärtner et al. [J Cereb Blood Flow Metab 1992;12:571-583] (MG PVEc), Rousset et al. [J Nucl Med 1998;39:904-911] (R PVEc), plus a modified version of MG PVEc (mMG PVEc) which uses the WM estimate from R PVEc.

The main program and GUI were written by NRU using Matlab, a programming tool widely diffused in the biomedical imaging. The software allows to apply a set of labels, defined a priori in the Talairach space to the segmented co-registered MRI [NeuroImage 2002;17:373-384]. Resulting ROIs are then transferred onto the PET/SPET study, and corresponding values are corrected according to R PVEc, as well as onto the images obtained by M PVEc, MG PVEc and mMG PVEc, providing corresponding corrected values. The software also includes:

- An original PVE-correction technique achieved by IBB.
- An original tool for fully automated definition of PVE-corrected white matter value.
- A fully automated Volume of Interest definition method, which now allows also a fully automated definition, when a normalisation matrix is available from external software (e.g. SPM).

The results of the validation are reported in the paper by Quarantelli et al., entitled "An integrated software for the analysis of brain PET/SPECT studies with Partial Volume Effect Correction", published by the Journal of Nuclear Medicine (J Nucl Med. 2004 Feb; 45(2): 192-201). In this work accuracy and precision of these methods have been measured applying the software to four simulated FDG-PET studies and subsequently introducing larger experimental errors, including co-registration errors (0 to 6 pixel mis-registration), segmentation errors (-13.7% to +14.1% GM volume change) and resolution estimate errors (-16.9% to 26.8% FWHM mismatch). Even in absence of segmentation and co-registration errors, the uncorrected PET values showed -37.6% GM underestimation and 91.7% WM overestimation. M PVEc left a residual underestimation of GM values (-21.2%).

Application of R PVEc and mMG PVEc provided accuracy above 96%. Software documentation has been included in SW package, and is being upgraded according to feedback from end-users. The software has been tested through fully and is in use in XX different research centres selected as Beta test sites. Research centres can ask to download the software to act as beta sites (, password required, send email to PVEOut consortium adopted the policy of create some beta cites in research centres to increase the initial core of users thus increasing the industrial interest of the PVEOut results.

Información relacionada

Reported by

Neurobiology Research Unit, Rigshospitalet, N9201
9 Blegdamsvej 9
2100 Koepenhagen
See on map
Síganos en: RSS Facebook Twitter YouTube Gestionado por la Oficina de Publicaciones de la UE Arriba