Abstract Archives of the RSNA, 2011


Virtual Childbirth for Mechanical Risk Assessment and Prevention of Dystocia

Scientific Formal (Paper) Presentations

Presented on November 29, 2011
Presented as part of SSJ12: Informatics (New Technologies)


Olivier Ami MD, PhD, Presenter: Nothing to Disclose
Lucie Cassagnes MD, Abstract Co-Author: Nothing to Disclose
Jean-Francois Uhl MD, Abstract Co-Author: Research partner, NVIDIA Corporation Research partner, Usefulprogress
Didier Lemery MD, PhD, Abstract Co-Author: Nothing to Disclose
Vincent D Delmas, Abstract Co-Author: Nothing to Disclose
Gerard Mage MD, PhD, Abstract Co-Author: Nothing to Disclose
Louis Bernard Boyer MD, Abstract Co-Author: Nothing to Disclose


Childbirth simulation is a new and promising tool to detect labor dystocia and seem to be a significant improvement over pelvimetry. Further studies are needed to confirm these preliminary results.


The aim of this study was to evaluate the diagnostic interest of vaginal delivery simulation from twenty four dimensional vectorial reconstruction of the maternal pelvis and fetus with magnetic resonance imaging. 24 pregnant women who underwent a pelvimetry with MRI for clinical and/or sonographical suspicion of cephalopelvic disproportion. Both fetus and pelvis were manually segmented for 3D vectorial reconstruction with Pelvitest (Ogami - France) software. Favorability of childbirth was estimated using the percentage of possibility to deliver regarding tested possibilities for fetal engagement, descent and rotation for each variety of presentation tested. Childbirth was defined at high risk of dystocia when the score of favourability was less than 25 %, mildly favourable between 25 and 50 %, and highly favourable when the score was greater than 50 %.


Of the 24 pregnant women tested, 13 delivered normally with an estimated highly favourable score by the simulator. 3 were delivered by elective cesarean section (2 presented with severe macrosomia and a score estimating them at high risk of dystocia, and 1 presented with small pelvis and a score estimating the score mildly favourable). 5 were delivered by emergency cesarean section (2 for cardiac fetal rythm abnormalities of which one had a mildly favourable score and the other one presented with a favourable score, and 3 for obstruction of labour with an estimated score at high risk of dystocia). 3 patients were delivered with a vacuum extraction and presented with a mildly favourable score. Considering the cases where biomechanical predictions of favourability could be verified, the simulation predictions were significantly coherent with the reality of childbirth (p < 0,0001).


Collision detection was performed between the fetal head and the maternal pelvis using selected possibilities of head's flexion, rotation, and synclitism.

Cite This Abstract

Ami, O, Cassagnes, L, Uhl, J, Lemery, D, Delmas, V, Mage, G, Boyer, L, Virtual Childbirth for Mechanical Risk Assessment and Prevention of Dystocia.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11016001.html