Authors: G. Wang and M. Wang
Affilation: Georgia Institute of Technology and Emory University, United States
Pages: 229 - 231
Keywords: High Performance Computing, Quantum Dot, GPU, Cancer Detection, Therapeutics
Bionanotechnologies such as quantum dots (QDs) or molecular beacons have generated huge amount of nanoimaging data to study in vitro and in vivo cell behavior for disease early diagnosis and therapeutic drug target. These images usually contain various cell or sub-cellular environment highlighted by multi-color clustered molecular beacons or thousands of single particle quantum dots. Quantifying the cell state or tracking biomarkers/therapeutic-targets requires robust segmentation of cell morphology in 2-D or 3-D, colocalization detection, and 1-D, 2-D, and 3-D multiple dot tracking. To accomplish the goal, the Bio-MIBLab (Bio-medical Inforamtics and Bioimaging Lab) at Biomedical Engineering has conducted extensive research in mathematical methods, computational software, and hardware methods to analyze high volume of bionanoimaging data with accuracy and efficiency. First, we have researched topology-independent image analysis methods, Gaussian fitting, and robust Kalman tracking to track 100s of bio-conjugated QDs for cancer early detecion. Second, we have researched advanced hardware approach such as programmable graphics processing units (GPUs). Third, we have researched to utilize a state-of-art 68-CPU Itanium2 linux cluster, which is donated to our lab by Hewlett Packard to research parallel tracking algorithm. All these computational methodologies provide high-speed, high-accuracy quantification for bionanotechnologies for cancer research.
Nanotech Conference Proceedings are now published in the TechConnect Briefs