Automated TMA-Core-Detection Algorithm

Abstract

Tissue microarray (TMA) is a high-throughput technology for the analysis of molecular markers in oncology. This method supports the presentation of several different tissue samples -TMA cores- in one singe glass slide. However, because of the large size of TMA cores, the “identification and analysis” procedure is a more or less time-consuming method. The TMA core-finding algorithm detailed in this study detects each of the TMA cores on the slide and it creates outline annotation around the cores automatically. A validation study is also presented, through which detection accuracy of this algorithm for detecting cores on brightfield and fluorescent slides have been measured. We have found a 77.5% detection accuracy in average, so based on this result we can conclude that our TMA core detection solution can be utilized as a useful tool for supporting TMA analysis.