Cancer remains a persistent global health challenge, with approximately 9.7 million deaths in 2022 (World Health Organization), even with expanding therapeutic advancements. Two‐dimensional (2D) cell culture models are fundamental to oncology research, although they fail to mimic the complexity of tumor micro‐environments. Three‐dimensional (3D) in vitro models such as spheroids can more precisely replicate drug responses. However, conventional methods of spheroid generation and analysis are subject to inconsistent outcomes and subjective interpretation, limiting their clinical translation potential. The integration of robotics and artificial intelligence (AI) for generating and analyzing in vitro models could improve the standardization and increase throughput. This study presents a framework that combines robot‐assisted spheroid generation via the robotic‐enabled biological automation (ReBiA) system with AI‐based characterization through deep learning models (ResNet50 for classification and YOLOv8 for segmentation). To showcase the capability of the system, non‐small cell lung cancer (NSCLC) and pancreatic ductal adenocarcinoma (PDAC) spheroids were generated using the robot‐assisted platform. Additionally, drug combinations of JQ1 and GANT61, planned through statistical experimental design, were applied to PDAC spheroids with drug response quantified via AI‐powered image analysis, to show the potential of the framework for efficient drug screening towards personalized treatment approaches.
Mahdy et al. (2026) studied this question.