Dominik Waibel

Label-free imaging of 3D pluripotent stem cell differentiation dynamics on chip

The paper introduces a label-free deep learning method to predict nuclear staining in 3D stem cell cultures from bright-field images, enabling high-resolution tracking of differentiation dynamics.

PXPermute: Unveiling staining importance in multichannel fluorescence microscopy

PXPermute identifies important fluorescent channels in IFC data by permuting pixel values and evaluating model performance, showing superior accuracy in channel selection.

SHAPR predicts 3D cell shapes from 2D microscopic images

SHAPR predicts 3D cell shapes from 2D images, improving accuracy over existing methods and enabling morphometric analysis of cell aggregates.

InstantDL - An easy-to-use deep learning pipeline for image segmentation and classification

InstantDL provides a deep learning pipeline for biomedical image tasks with minimal effort, achieving competitive performance across seven datasets without parameter tuning.

Systematic comparison of incomplete-supervision approaches for biomedical imaging classification

The paper compares incomplete-supervision methods for biomedical image classification, finding that pre-training and semi-supervised learning contribute more than active learning to performance.