Fast cell detection in high-throughput imagery using GPU-accelerated machine learning
Beckman Institute for Advanced Science and Technology, University of Illinois Urbana-Champaign, USA
IEEE International Symposium on Biomedical Imaging: From Nano to Macro, 2011
@inproceedings{mayerich2011fast,
title={Fast cell detection in high-throughput imagery using GPU-accelerated machine learning},
author={Mayerich, D. and Kwon, J. and Panchal, A. and Keyser, J. and Choe, Y.},
booktitle={Biomedical Imaging: From Nano to Macro, 2011 IEEE International Symposium on},
pages={719–723},
year={2011},
organization={IEEE}
}
High-throughput microscopy allows fast imaging of large tissue samples, producing an unprecedented amount of sub-cellular information. The size and complexity of these data sets often out-scale current reconstruction algorithms. Overcoming this computational bottleneck requires extensive parallel processing and scalable algorithms. As high-throughput imaging techniques move into main stream research, processing must also be inexpensive and easily available. In this paper, we describe a method for cell soma detection in Knife-Edge Scanning Microscopy (KESM) using machine learning. The proposed method requires very little training data and can be mapped to consumer graphics hardware, allowing us to perform real-time cell detection at a rate that exceeds the data rate of KESM.
June 24, 2011 by hgpu
