An algorithm for quantifying filopodia of multiple cells with weak fluorescence signal

Filopodia function as cellular sensors, detecting the microenvironment and directing cell migration. They play a crucial role in cancer metastasis. Quantifying the filopodia characteristics of cancer cells is a prerequisite for studying the complex role of filopodia in cancer cell metastasis. Several algorithms have been developed, yet most of these algorithms are typically suited for extracting filopodia from individual cells. This paper aims to develop an independent algorithm (MC-FiloAssay) for quantifying filopodia in multi-cell environments. The filopodia of nasopharyngeal carcinoma cells (CNE2 and 5–8F) and normal nasopharyngeal epithelial cells (NP69) were quantified with MC-FiloAssay. A linear regression analysis comparing filopodia lengths measured by MC-FiloAssay and manual annotation yielded a coefficient of determination ([Formula: see text]), indicating high accuracy in multi-cell filopodia extraction. Furthermore, MC-FiloAssay outperforms existing algorithms under low signal conditions and in multi-cell fields of view. Analysis of CNE2 cells at different confluences revealed that confluence does not affect filopodia length or width but influences filopodia density. Additionally, significant differences were observed between CNE2 and the other two cell lines (5–8 F and NP69): CNE2 filopodia were longer, thinner, and more densely distributed. These results demonstrate that MC-FiloAssay is a robust tool for multi-cell filopodia quantification.