Circulating tumor cell detection in cancer patients using in-flow deep learning holography

Area of Science:

• Biomedical optics
• Microfluidics
• Artificial intelligence in medicine
• Cancer diagnostics

Background:

• Circulating tumor cells (CTCs) are vital biomarkers for cancer detection and monitoring, but their low abundance and heterogeneity pose significant detection challenges.
• Traditional imaging methods for CTCs lack the sensitivity and specificity required for accurate enumeration, often missing crucial cell populations.

Purpose of the Study:

• To develop and validate a streamlined Digital Holographic Microscopy (DHM)-based system for high-throughput, label-free CTC identification.
• To improve the sensitivity and specificity of CTC enumeration by integrating microfluidic enrichment, deep learning image analysis, and immunofluorescent profiling.

Main Methods:

• A DHM system integrated with inertial microfluidics for cell enrichment and dual-modality imaging (holography and fluorescence sensing).
• A deep learning model trained on diverse blood samples and cancer cell lines for real-time morphological cell analysis.
• Immunofluorescent profiling, including markers beyond EpCAM, to enhance CTC identification accuracy.

Main Results:

• The DHM platform demonstrated higher CTC counts in late-stage prostate cancer patients compared to healthy controls, with a low false positive rate (1 cell/mL).
• A significant proportion of identified CTCs were EpCAM-negative but PSMA-positive, highlighting limitations of EpCAM-based detection.
• The system provides a morphological confidence score per cell, combinable with immunofluorescence for precise CTC enumeration.

Conclusions:

• The integrated DHM and deep learning system offers a powerful, label-free approach for sensitive and specific CTC detection.
• This technology has the potential to overcome limitations of traditional CTC detection methods, particularly for EpCAM-negative CTCs.
• The platform shows promise for applications in cancer screening, diagnostics, prognostication, and precision oncology.