Researchers at Taiyuan University of Technology and Shanxi Bethune Hospital have combined an inertial microfluidic chip with a deep learning model to create a label-free platform that isolates and identifies circulating tumor cells (CTCs). They reported the work in Biomedical Analysis.
CTCs detach from tumors and enter the bloodstream, where they can contribute to cancer metastasis. As a result, they may serve as useful biomarkers for cancer diagnosis and treatment monitoring. However, isolating and precisely identifying CTCs remains challenging because they are extremely rare—approximately one to ten cells per billion blood cells—and are surrounded by a complex mixture of other blood cells.
Many microfluidic methods use antibodies to capture CTCs by targeting specific surface markers, but they may miss cells that lack those markers. Label-free methods separate cells based on physical characteristics such as size. Inertial microfluidic chips, for example, are relatively straightforward, low-cost, and high-throughput. However, researchers often still need immunolabeling to identify the sorted cells, adding steps and potentially compromising cell viability.
CTCs are generally larger ...