Dongfeng H. Email & Phone Number

Greater Pittsburgh Area

Machine Learning for Sudden Cardiac Arrest (SCA) detection and prediction algorithms for business applications.

Greater Minneapolis-St. Paul Area

Machine learning (deep learning) for cardiac arrhythmia classification.System designs and algorithms in The EnSite Precision™ cardiac mapping system for cardiac arrhythmia treatment in electrophysiology (EP), such as activation detection, ECG analysis, advanced cardiac activation map, etc.C++, Python, Linux.

Princeton, Nj

Working on algorithm design & implementation for image-guided navigation system. Clinical workflow design and implementation. Fast deformable 3D medical image registration for surgery navigation system. Fast 3D medical image segmentation for surgery navigation system. Tools: C/C++, Various data structures & algorithms; GPU, Multi-Thread, ITK, VTK, QT, VS, OpenCV, OpenGL, Boost.

Developed novelty computational tools improving the efficacy of 4-D image-guided radiation therapy by accounting for respiratory motion during the treatmentAdvanced a novel method to automatically identify artifacts during 4-D CT images and proposed a novel graph-based algorithm to reduce the magnitude of artifacts present in helical 4-D CT images by structure alignment. Comprehensively quantify the frequency and magnitude of helical 4-D CT artifacts on a large data set from 81 patients with total 139073 slices and total 1264 respiratory periods which corresponds to total 12640 image stacksProposed novel algorithms on graph search with application on image/video resizing and medical image segmentation

Proposed a novel technique of learning model with spatial and statistical information for semantic image classification that can captured the spatial relations. Comprehensively evaluated existing methods on local interest points detection and matchingDeveloped a fast interactive image segmentation algorithm based on graph cut minimization techniquePresented a three dimensional reconstruction method based on image sequence of multiple images

Proposed a novel image classification algorithm based on random semi-supervised sampling which can use both labeled and unlabelled dataComprehensively evaluated existing experimental methods on image classification and demonstrated the necessity for developing technology in new directions

Postdoc, Medical Image Analysis

Ph.D., Computer Science And Technology

Activities and Societies: Image Process and Pattern RecongitionGraduated with honour Published over 30 high quality papers in the computer.

M.S., Computer Science And Technology

Activities and Societies: Image Process

B.E., Computer Science And Technology