Non-Invasive Optical Imaging in vivo for Early Detection and Advanced Diagnosis of Cancer - Robotics Institute Carnegie Mellon University

We are exploring cutting edge optical imaging technology combined with robotics and computer vision technology and statistical learning algorithms to find the most discriminative image features for cancer detection and diagnosis. This work is carried out by an interdisciplinary team of researchers from both Carnegie Mellon University and University of Pittsburgh, and is supported by the unconventional innovation program of National Cancer Institute.

The research focus in the project is twofold:

develop methodology for image feature extraction from multispectral biological images, find the most discriminating feature subsets, achieve high classification rates with minimum false positive rate, and provide direct feedback to the imaging process.
construct a 3D robotics imaging system using multiple cameras and lighting sources for detection of skin cancer through reconstruction of 4D spatiotemporal images.

Displaying 7 Publications

2005

SVM Decision Boundary Based Discriminative Subspace Induction

Jiayong Zhang and Yanxi Liu

Journal Article, Pattern Recognition, Vol. 38, No. 10, pp. 1746 - 1758, October, 2005

2004

Cervical Cancer Detection Using SVM Based Feature Screening

Jiayong Zhang and Yanxi Liu

Conference Paper, Proceedings of 7th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI '04), pp. 873 - 880, September, 2004

2002

Does Multispectral Texture Features Really Improve Cervical Cancer Detection?

Tong Zhao, Jiayong Zhang, and Yanxi Liu

Conference Paper, Proceedings of International Conference on Diagnostic Imaging and Analysis (ICDIA '02), August, 2002

SVM Based Feature Screening Applied to Hierarchical Cervial Cancer Detection

Jiayong Zhang, Yanxi Liu, and Tong Zhao

Conference Paper, Proceedings of International Conference on Diagnostic Imaging and Analysis (ICDIA '02), August, 2002

Learning Multispectral Texture Features for Cervical Cancer Detection

Yanxi Liu, Tong Zhao, and Jiayong Zhang

Conference Paper, Proceedings of IEEE International Symposium on Biomedical Imaging (ISBI '02), pp. 169 - 172, July, 2002

SVM Decision Boundary Based Discriminative Subspace Induction

Jiayong Zhang and Yanxi Liu