FEtus and MOther NUmerical Models for Dosimetry Studies
Abstract
——————————————————-
This project aims at creating models of fetus and pregnant woman for dosimetry studies which are of crucial needs, as acknowledged by the WHO and the COST 281. NumeriCé models of mothers and fetus, evolving along the course of pregnancy, and positioned intl different spatial configurations, will be developed, based on segmentation of medical images. These models and their posing methods will be made available to the scientific community.
——————————————————-
Partners
ENST – Laboratoire Traitement et Communication de l’Information
UMR CNRS 5141
Hôpital Saint Vincent de Paul – AP-HP
Philips Medical Systems Research Paris
Institut Broca-Sainte Anne, INSERM U796 http://www.broca.inserm.fr/ifr77/index.html
Contact
Elsa ANGELINI ENST – LTCI
Duration
24 months
Summary
——————————————————-
1) Introduction
Fields of application: Several studies have been carried out on adult and children heads te develop numerical models, used for RF wave propagation simulations and dosimetry studies However, two important questions are currently raised: studies of fetus and pregnanl women are needed, and models should be able to undergo deformations in order to simulate different positions of the body. This project aims at generating mother/fetus models vic integration of segmentation results obtained on medical images, at different gestationa stages and in different positions. Models will be made available to the scientific community.
Keywords : 3D models of human beings in situation, fetus, tools for dosimetry studies.
2) State of the art
Few models of pregnant women and fetus have been developed, either in stylized form, based on computer assisted conception, or in voxelized form, based on segmentation of post-mortem or incomplete image data. This field of research remains wide open. Innovative aspects of our project include the use of in vivo and anatomically correct images, acquired at different stages of development, and the development of methodological approaches to deform our models with physiological constraints, and simulate different body positions.
ln addition to methodological innovations for segmentation of 3D ultrasound and MRI prenatal images, for which very little work exists, this project targets a con crete outcome of primary importance, providing dosimetry research teams with realistic models and contributing to public health education and information.
3) Scientific methodology
The scientific methodology includes:
1- Segmentation of 3D echographic images of 1 sI and 3rd trimesters, to build a fetus modal.
2- Segmentation of prenatal MRI to extract abdominal tissues of the mother (skin, bones, fat, muscles) and internai organs of the fetus (brain, lungs, heart).
3- Fusion and registration of MRI and ultrasound segmentation results in a hybrid CAD/voxelized mother/fetus modal.
4- Simulation of deformations of the fetus to modify its intra-uterine position and of the mother in different positions.
5- Morphing of the model to simulate the growth of the fetus during pregnancy.
6- Clinical validation of the models from the neuropediatricians and obstetricians, partners of the project.
4) Partners, skills
The work planned in this project will be performed at ENST, CNRS UMR 5141, which has recognized skills in medical image analysis, 3D segmentation and modeling, and benefits from a previous experience on two projects for adult and children head modeling for the study of biological effects of RF waves. Availability of a large database of medical image data is guaranteed by collaborations with Philips Medical Systems Research Paris for ultrasound images and Saint-Vincent de Paul hospital for prenatal MRI data.