Research

We study the problem of 3D object generation. We propose a novel framework, 3D Generative Adversarial Network (3D-GAN), leveraging recent advances in volumetric convolutional networks and generative adversarial nets.

We develop a computational model that explains how people make causal judgments in physical scenes by mentally simulating counterfactual outcomes and comparing those to what actually happened.

In order to be able to design synthetic organs that function autonomously, we will need to engineer artificial tissue homeostasis. To control the size of these artificial tissues, two major mechanisms will have to be engineered.

We develop image analysis algorithms for whole-heart cardiac MRI in patients with severe congenital heart disease.

Integrating multi-modal clinical data and using recurrent and convolution neural networks to predict when patients will need important interventions.

We develop novel algorithms for neuroimage segmentation, registration, and joint modeling of brain structure with clinical and genetic data.

Knitting is the new 3d printing. It has become popular again with the widespread availability of patterns and templates, together with the maker movements. Lower-cost industrial knitting machines are starting to emerge, but we are still missing the corresponding design tools. Our goal is to fill this gap.

A system that can help you to design our own drone!

Our goal is to investigate deterministic algorithms for robotic task and motion planning.

We use computational models of biological pathways in disease to estimate the range of possible responses to various hypothetical drugs, given patient variability and limited pathway-relevant data.

Existing methods for cloning and recombination of DNA enable construction of arbitrary sequences. However, the sequential nature of these techniques makes them time-consuming and expensive. Furthermore, while the transformation of an existing plasmid into a host strain can be reliable when a selection marker is used, there are many current limitations: the number of different plasmids that can be co-transformed is limited by the choice of markers and compatible origins of replication; plasmids are less stable than chromosomal DNA and are difficult to maintain indefinitely without mutation; and cistronic interactions cannot be designed since each new nucleotide sequence added is on an unconnected DNA molecule. To overcome these limitations, we are designing reconfigurable chromosomes consisting of both fixed and variable regions. While the fixed region is carefully optimized and tuned ahead of time, the variable region can be modified in the field, at the point-of-use, leading to rapid and on-demand realization of novel biocircuits with many different phenotypes.

A new approach for designs in multi-material fabrication

Developing state-of-the-art tools that process 3D surfaces and volumes

Help robots learn faster by providing demonstrations when they need help

We aim to study the causes and transmission modes of infectious diseases among members of a community in the presence of hidden, asymptomatic spreaders of the pathogen.

Almost every object we use is developed with computer-aided design (CAD). While CAD programs are good for creating designs, using them to actually improve existing designs can be difficult and time-consuming.

Our goal is to develop better metagenomic binning by origin species of fragments of sequenced environmental DNA.

We develop algorithms for fetal MRI interpretation, to enable noninvasive fetal monitoring from MRI.

We develop, parameterize, and validate a model for tumor growth inhibition using in vivo mouse data and study the effects of modeling uncertainty and inter-individual variability on drug candidate efficacy predictions.

Transitioning machine learning models across electronic health record (EHR) versions can be improved by mapping different EHR encodings to a common vocabulary.

We develop machine learning algorithms to automatically and quantitatively assess the severity of pulmonary edema from chest x-ray images.

We develop algorithms, systems and software architectures for automating reconstruction of accurate representations of neural tissue structures, such as nanometer-scale neurons' morphology and synaptic connections in the mammalian cortex.

Using adversarial signals and a cycle-consistency based regularization, we can supplement paired regression tasks with unpaired data to improve regression performance.

Developing state-of-the-art deep learning algorithms for analyzing and modeling 3D geometry