Artificial intelligence-driven morphology-based enrichment of malignant cells from body fluid

• Published in: Modern pathology Vol. 36; no. 8; p. 100195
• Main Authors: Mavropoulos, Anastasia, Johnson, Chassidy, Lu, Vivian, Nieto, Jordan, Schneider, Emilie C., Saini, Kiran, Phelan, Michael L., Hsie, Linda X., Wang, Maggie J., Cruz, Janifer, Mei, Jeanette, Kim, Julie J., Lian, Zhouyang, Li, Nianzhen, Boutet, Stephane C., Wong-Thai, Amy Y., Yu, Weibo, Lu, Qing-Yi, Kim, Teresa, Geng, Yipeng, Masaeli, Maddison (Mahdokht), Lee, Thomas D., Rao, Jianyu
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.08.2023
• Subjects: carcinoma; cytology; effusion; enrichment; machine learning; morphology; cytology; enrichment; effusion; morphology; machine learning; carcinoma
• ISSN: 0893-3952; 1530-0285
• DOI: 10.1016/j.modpat.2023.100195

Cell morphology is a fundamental feature used to evaluate patient specimens in pathologic analysis. However, traditional cytopathology analysis of patient effusion samples is limited by low tumor cell abundance coupled with the high background of nonmalignant cells, restricting the ability of downstream molecular and functional analyses to identify actionable therapeutic targets. We applied the Deepcell platform that combines microfluidic sorting, brightfield imaging, and real-time deep learning interpretations based on multidimensional morphology to enrich carcinoma cells from malignant effusions without cell staining or labels. Carcinoma cell enrichment was validated with whole genome sequencing and targeted mutation analysis, which showed a higher sensitivity for detection of tumor fractions and critical somatic variant mutations that were initially at low levels or undetectable in presort patient samples. Our study demonstrates the feasibility and added value of supplementing traditional morphology-based cytology with deep learning, multidimensional morphology analysis, and microfluidic sorting.