| Issue |
Rev Orthop Dento Faciale
Volume 30, Number 3, Septembre 1996
|
|
|---|---|---|
| Page(s) | 369 - 379 | |
| DOI | https://doi.org/10.1051/odf/1996029 | |
| Published online | 30 mars 2010 | |
Les réseaux de neurones et l'étude de l'architecture faciale
1 a, avenue de Haguenau, 67300 Schiltigheim
Résumé
Le concept de réseaux de neurones appliqué à l'étude de l'architecture faciale a permis de mettre en évidence des relations très étroites entre la forme de la calvaria et certaines valeurs faciales, en particulier le prognathisme des deux mâchoires.
L'existence de ces relations permet de définir pour un crâne donné une face “normale”. Il s'agit là d'une individualisation poussée du diagnostic puisque la référence de normalité n'est plus externe au patient (moyenne ou table) mais interne.
Abstract
The so-called “neural networks” computer programs have been developped to try to mimic the human brain activity. In this paper the neural network concept is applied to the study of facial architecture. Relationships between the shape of the calvarium and facial measurements are shown. These relationships are especially strong with the maxillary and the mandibular prognathisms.
The antero-posterior position of the face relative to the cranial base appears to be closely related to the cranial base itself and to the calvarium.
The beeing of these relationships makes possible to define, for a given skull, a “normal” face and a normal position of this face related to the cranial base.
This is a further step in the way of individualizing the diagnosis. The reference for normality is not external to the patient (mean or table) but inbuilt.
Mots clés : Réseaux de neurones / Morphologie / Diagnostic
Key words: Neural networks / Morphology / Diagnosis
© Revue d'Orthopédie Dento-Faciale, Paris, 1996
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.

