Published in Scientific Reports, the SCOUT pipeline enables rapid staining, imaging, and analysis of 3D cerebral organoids. We applied SCOUT to compare different culture protocols as well as to study how Zika virus inhibits organoid growth.

SCOUT's computational pipeline is a Python package with a pre-built Docker image. The package also includes a pre-trained U-Net model for segmentation of ventricular zones in cerebral organoids based solely on nuclear staining.

Awarded 1st place at the 2019 MIT Machine Learning Across Disciplines Challenge, the Samply sample manager organizes audio samples using machine learning.

The sample manager uses a convolutional neural network (CNN) and a parametric approximation to the t-distributed stochastic neighbor embedding (t-SNE) algorithm to generate visualizations of audio sample libraries. This work was later presented at the 2020 Audio Engineering Society conference.