Ultrasound-Based Assessment of Brain Folding Patterns in Early Pregnancy

Normal development of the human brain is characterized by precisely timed growth and folding of its surface (cortex).

This dynamic process provides a phenotypic report of overall brain development, and deviations are associated with poor cognitive outcomes.

Advances in ultrasound (US) imaging technology now make it possible to visualize the cortex and screen for brain abnormalities before birth, from as early as 14 gestational weeks (GW).

However, diagnostic confirmation currently relies on expensive, specialized magnetic resonance (MR) imaging, following referral from an initial US scan.

Referral procedures delay diagnosis until 20 GW (at the earliest), and limit detailed evaluation to a minority of fetuses.

Given that US is a first step in the pregnancy care continuum, dedicated US-based tools would transform diagnosis from specialist care to standard, routine scanning.

We propose to develop image analysis algorithms to quantify brain folding (and possible deviations) from standard US data. We will develop a surface-based population atlas of the fetal brain surface for each week of gestation.

With images from 4000+ healthy fetuses, our surface atlas will depict the brain’s constantly evolving topology, representing gestational development from 6 weeks earlier than previously possible with MR imaging.

We will use advanced machine learning-based techniques to extract features from the US signal, and quantify geometric characteristics of each fold.

Our population atlas will be a normative reference against which to compare an individual’s brain development, with potential for detection and tracking of morphological abnormalities from the earliest prenatal scan, reducing reliance on secondary MR assessment.