Wentao Zhu

This paper introduces DeepIso, a self-supervised framework that enhances isotropic resolution in 3D fluorescence microscopy by restoring axial frequency content without requiring explicit PSF models or system calibration.

This paper introduces an algorithm to reconstruct 3D molecular orientation and density from polarized fluorescence microscopy data, demonstrating its application on biological structures and dynamics.

AnatomyNet is a 3D squeeze-and-excitation U-Net for fast and automated whole-volume head and neck anatomical segmentation, achieving improved Dice similarity coefficients over existing methods.

DeepEM improves pulmonary nodule detection in CT images by using weakly supervised labels from electronic medical records.

On LIDC-IDRI, gradient boosting on 3D dual path network features reaches 90.44% accuracy on annotated nodules; with a 3D Faster R-CNN detector, patient-level accuracy is 81.41%, 99% of doctors' average.

This paper proposes deep multi-instance networks for whole mammogram classification without requiring manual annotations, showing robustness on the INbreast dataset.

An end-to-end fully convolutional network with a positional prior, conditional random field structural learning, and adversarial training segments mammographic masses on INbreast and DDSM-BCRP.