SU-E-T-253: Open-Source Automatic Software for Quantifying Biological Assays of Radiation Effects




Clonogenic cell survival is a common assay for quantifying the effect of drugs and radiation. Manual counting of surviving colonies can take 30–90seconds per plate, a major limitation for large studies. Currently available automatic counting tools are not easily modified for radiation oncology research. Our goal is to provide an open-source toolkit for precise, accurate and fast analysis of biological assays in radiation oncology.


As an example analysis, we used HeLa cells incubated with gadolinium nanoparticles prior to irradiation. After treatment, the cells are grown for 14days to allow for colony formation. To analyze the colony growth, we capture images of each dish for archiving and automatic computer-based analysis. A FujifilmX20 camera is placed at the top of a box setup, 20cm above the sample, which is backlit by a LED lamp placed at the bottom of the box. We use a Gaussian filter (width=1.3mm) and color threshold (19–255). The minimum size for a colony to be counted is 1mm. For this example, 20 dishes with a large range of colonies were analyzed. Each dish was counted 3 times manually by 3 different users and then compared to our counter.


Automatic counting of cell colonies takes an average of 7seconds, enabling the analysis process to be accelerated 4–12 times. The average precision of the automatic counter was 1.7%. The Student t-test demonstrated the non-significant differences between the two counting methods (p=0.64). The ICC demonstrated the reliability of each method with ICC>0.999 (automatic) and ICC=0.95 (manual).


We developed an open-source automatic toolkit for the analysis of biological assays in radiation oncology and demonstrated the accuracy, precision and effort savings for clonogenic cell survival quantification. This toolkit is currently being used in two laboratories for routine experimental analysis and will be made freely available on our departmental website.