Dahle Jostein, Kakar Manish, Steen Harald B, Kaalhus Olav
Department of Radiation Biology, Norwegian Radium Hospital, Montebello, Oslo, Norway.
Cytometry A. 2004 Aug;60(2):182-8. doi: 10.1002/cyto.a.20038.
Clonogenic assays are used frequently to measure the cell killing and mutagenic effects of radiation and other agents. Clonogenic assays carried out manually are tedious and time-consuming and involve a significant element of subjectivity. However, several commercial automatic colony counters are available. Based on CCD video imaging and image analysis they are relatively expensive and can analyze only one petri dish at a time.
We have developed a cheaper and more efficient device, which employs a flat bed scanner to image 12 60-mm petri dishes at a time. Two major problems in automated colony counting are the clustering of colonies and edge effects. By using standard image analysis and implementing an inflection point algorithm, these problems were greatly diminished. The resulting system was compared with two manual colony counts, as well as with automated counts with the Oxford Optronix ColCount colony counter for cell lines V79 and HaCaT.
Comparisons assuming the manual counts to be correct showed that our automatic counter was slightly more accurate than the commercial unit.
As a whole, our automated colony counter performed significantly better than the commercial unit with regard to processing time, cost and accuracy.
克隆形成试验常用于测量辐射及其他因素对细胞的杀伤和诱变作用。手动进行克隆形成试验既繁琐又耗时,且存在较大主观因素。不过,市面上有几款商用自动菌落计数器。基于电荷耦合器件(CCD)视频成像和图像分析技术,它们相对昂贵,且每次只能分析一个培养皿。
我们研发了一种更便宜且高效的设备,它利用平板扫描仪一次对12个60毫米培养皿进行成像。自动菌落计数中的两个主要问题是菌落聚集和边缘效应。通过使用标准图像分析并实施拐点算法,这些问题得到了极大缓解。将所得系统与两次手动菌落计数结果进行比较,同时也与使用牛津奥普特尼克斯公司的ColCount菌落计数器对V79和HaCaT细胞系进行的自动计数结果作比较。
假设手动计数正确进行比较,结果显示我们的自动计数器比商用设备稍精确。
总体而言,我们的自动菌落计数器在处理时间、成本和准确性方面比商用设备表现更佳。