As a constitutent component of CIFASD4, this project will focus on earlier detection of the neurofacial effects of
prenatal alcohol exposure. Four topics of the Request for Application (RFA) will be targeted: a) improved
FAS/FASD facial recognition through 3D photography and computer analyses in individuals of different age
groups; b) the utility of prenatal ultrasound as a screening or diagnostic modality; c) 3D facial imaging during
the neonatal period to detect more subtle facial features affected by prenatal alcohol exposure; d) innovative
uses of technologies, including handheld devices with apps, to screen for dysmorphology.
Previously, we have implemented methods and associated software for analyzing postnatal face shape for the
detection of the effects of prenatal alcohol exposure. Although effective in their accuracy, they rely on the use
of expensive and relatively clumsy 3D cameras for which not insignificant training is required. New methods of
2D image analysis are available, mobile device acquisition of 3D images is imminent and online recruitment to
clinical research is becoming more common place. Finally, state-of-the-art machine learning methods are
proving effective in large-scale analysis of ultrasound and MRI data. To accomplish our goals, we propose the
following specific aims:
1. Automated screening of facial images for effects of prenatal alcohol exposure with potential for on-line and
mobile device use and integration of genetic, behavioral and cognitive data;
2. Fetal ultrasound analysis to detect facial, cranial and neural effects of prenatal alcohol exposure with
3. Algorithm and software development to improve current analysis of face-brain-alcohol interactions.
Achieving aim 1 will dramatically impact access to validated facial screening for prenatal alcohol exposure.
Successful demonstration of aim 2 will achieve the earliest possible diagnosis enabling further research on
interventions but also anticipatory neonatal management. Progress in aim 3 will enhance face-brain
morphometric analyses in collaboration with other CIFASD partners (U01:Parnell/Eberhart, U01:Wozniak).
We will work collaboratively with consortium partners who will recruit subjects and provide facial and
ultrasound images (U01: Chambers; U01: Coles; U01: Mattson; U01: Weinberg; U01: Wozniak). We will co-
operate on the analysis of images arising from basic science partners focusing on the use of animal models.
We will rely on research resources for validation of postnatal screening tools (R24 Dysmorphology - Jones)
and online/app development and data management (R24 Informatics - Barnett).