Automated Analysis of Liquid Biopsy Using Deep Learning: Detecting Circulating Tumor Cells and Cancer-Associated Fibroblasts.

Cell detection is one of the most significant tasks in circulating tumor cells (CTCs) and cancer-associated fibroblasts (CAFs) analysis, as these cells are emerging as potential biomarkers for cancer prognosis and diagnosis. Traditional approaches to cell detection are often manual, leading to long turnaround times and significant variability between experts. In this chapter, we first provide an overview of deep learning techniques for general cell detection, with a specific focus on CTC and CAF detection. We then discuss developments in optical imaging systems that offer high-resolution, high-quality, and all-in-focus imaging capabilities. By integrating advanced optical hardware with deep learning algorithms, we demonstrate high-accuracy and high-fidelity detection of CTCs and CAFs in microfiltered-based samples. We also emphasize the need for refined technologies and models to enhance the clinical utility of CTC characterization and deepen the understanding of metastasis. The integrated approach introduced in this chapter has the potential to establish a new automated paradigm for CTC and CAF analysis.