Quantitative analysis of living tumor cells using large-scale digital cell analysis system.

摘要

Study of living cells has been one of the important research endeavors. A new approach to quantitative analysis of live tumor cell image data is proposed; which can be applied to problems of general importance including cell pedigree analysis and cell tracking. By using time as an extra dimension, level set methods are employed to determine cell trajectories from 2D + time data sets. Cell cluster separation and mitotic cell detection steps are performed within the trajectories. Each of the separate cell trajectories corresponds to the motion pattern of an individual cell in the data set.; Compared with model-free methods, our method provides more accurate segmentation for closed-object contours, even for weak cell boundaries. Complete and accurate description of the cellular shape does not require extensive post-processing steps. Our approaches are also able to track multiple cells in high density sequences exhibiting cell-cell contacts.; Model-based techniques, such as active contours, produce closed and smooth object boundaries, and provide a first guess through the interactive initialization. This approach has a restriction on the image collection interval which will greatly affect the shape deformation or the displacement of the cells in adjacent frames. Our approach extracts the cell trajectory for long observation time.; Compared to pattern recognition based segmentation methods, our approach has a wider range of applications and can be easily transplanted to various cellular structure analysis objectives without a complicate rule definition and training.; The Large Scale Digital Cell Analysis System (LSDCAS) developed at the University of Iowa provides capabilities for live cell image acquisition. The developed method was tested on cancer cell image sequences and its performance compared with manually-defined ground truth. The similarity Kappa Index is 0.84 for segmentation area and the signed border positioning segmentation error is 1.69 +/- 2.17 mum.; The reported method is a substantial improvement over existing cell analysis approaches. The temporal information is fully utilized so that it yields more robust segmentation. The method applies the graph theory to describe the adjacency relationship which is a novel application for live cell analysis. Important information from cytology is used for mitosis detection and determination of cell count.