Cell image segmentation is a part of quantitative studies regarding cell movement and cell behavior, and it plays a critical role in molecular biology and cellular biochemistry. Therefore, it is fundamentally important to evaluate the performance levels of cell image segmentation algorithms. In our previous study, the performance metrics for cell image segmentation algorithms were proposed. The sampling variability can result in measurement uncertainties. In this article, the uncertainty of the measure, i.e., the total error rate, in the cell image segmentation is computed in terms of standard error and 95 % confidence interval using bootstrap method as well as an analytical method. Examples are provided.
Citation: NIST Interagency/Internal Report (NISTIR) - 7954Report Number:
NIST Pub Series: NIST Interagency/Internal Report (NISTIR)
Pub Type: NIST Pubs
Cell image segmentation, Misclassification error rate, Total error rate, Uncertainty, Standard error, Confidence interval, Bootstrap, Analytical method.