Radiomics on MRI for breast cancer risk stratification

We compute radiomic features on breast MRI to better characterize breast cancer-associated risks.

Supervised learning frameworks for MRI analysis

We implement deep learning frameworks for MRI analysis tasks involved in answering clinical questions, such as segmentation and synthesizing missing contrasts.

Untrained model-based reconstruction of multi-contrast MRI from highly accelerated raw data

We implemented a deep learning framework with physics based regularization for untrained, scan-specific reconstruction of multi-contrast MRI data.

Identifying and modelling tumor habitat growth on preclinical MRI

We identify three tumor habitats on preclinical, multimodal MRI of a murine model of glioblastoma and subsequently build math models to describe habitat growth.

Characterizing errors in perfusion modelling in DCE-MRI of breast cancer lesions

We investigated errors in perfusion model parameters induced by abbreviating DCE-MRI series in the breast.